Note that the paper-to-topic mappings are computed using an LLM and the topic mapper is itself a work in progress.
@inproceedings{Lotfi2024,
author = {Lotfi, F. and Virji, K. and Dudek, N. and Dudek, G.},
title = {A comparison of RL-based and PID controllers for 6-DOF swimming robots: hybrid underwater object tracking},
booktitle = {Proc Intelligent Robotics and Systems},
year = {2024},
note = {To appear},
url = {http://arxiv.org/abs/2401.16618},
eprint = {2401.16618},
archivePrefix = {arXiv},
primaryClass = {cs.RO},
month = {Jan},
year = {2024},
note = {Accessed: Apr. 11, 2024}
} abstract: In this paper, we present an exploration and assessment of employing a centralized deep Q-network (DQN) controller as a substitute for the prevalent use of PID controllers in the context of 6DOF swimming robots. Our primary focus centers on illustrating this transition with the specific case of underwater object tracking. DQN offers advantages such as data efficiency and off-policy learning, while remaining simpler to implement than other reinforcement learning methods. Given the absence of a dynamic model for our robot, we@inproceedings{Holliday2024,
author = {A. Holliday and G. Dudek},
title = {A Neural-Evolutionary Algorithm for Autonomous Transit Network Design},
booktitle = {Proc. International Conference on Robotics and Automation},
year = {2024},
month = {May},
url = {http://arxiv.org/abs/2403.07917}
} abstract: Planning a public transit network is a challenging optimization problem, but essential in order to realize the benefits of autonomous buses. We propose a novel algorithm for planning networks of routes for autonomous buses. We first train a graph neural net model as a policy for constructing route networks, and then use the policy as one of several mutation operators in a evolutionary algorithm. We evaluate this algorithm on a standard set of benchmarks for transit network design, and find that it outperforms the learned policy@inproceedings{konar2024accelerating,
author = {Abhisek Konar and Amal Feriani and Di Wu and Seowoo Jang and Xue Liu and Gregory Dudek},
title = {Accelerating Digital Twin Calibration with Warm-Start Bayesian Optimization},
booktitle = {Proc. ICC 2024-IEEE International Conference on Communications},
year = {2024},
pages = {2372--2377},
address = {Denver, CO, USA},
doi = {10.1109/ICC51166.2024.10622967}
} abstract: Digital twins are expected to play an important role in the widespread adaptation of AI-based networking solutions in the real world. The calibration of these virtual replicas is critical to ensure a trustworthy replication of the real environment. This work focuses on the input parameter calibration of radio access network (RAN) simulators using real network performance metrics as supervision signals. Usually, the RAN digital twin is considered a black-box function and each calibration problem is viewed as a standalone search problem@inproceedings{luo2024adaptive,
title={Adaptive dynamic programming for energy-efficient base station cell switching},
author={Luo, Junliang and Xu, Yi Tian and Wu, Di and Jenkin, Michael and Liu, Xue and Dudek, Gregory},
booktitle={2024 IEEE International Conference on Communications Workshops (ICC Workshops)},
pages={1365--1370},
year={2024},
organization={IEEE}
} abstract: Energy saving in wireless networks is growing in importance due to increasing demand for evolving new-gen cellular networks, environmental and regulatory concerns, and potential energy crises arising from geopolitical tensions. In this work, we propose an approximate dynamic programming (ADP)-based method coupled with online optimization to switch on/off the cells of base stations to reduce network power consumption while maintaining adequate Quality of Service (QoS) metrics. We use a multilayer perceptron (MLP) given@inproceedings{li2024anomaly,
title={Anomaly Detection for Scalable Task Grouping in Reinforcement Learning-based RAN Optimization},
author={Li, Jimmy and Kozlov, Igor and Wu, Di and Liu, Xue and Dudek, Gregory},
booktitle={2024 IEEE International Conference on Communications Workshops (ICC Workshops)},
pages={1395--1400},
year={2024},
organization={IEEE}
} abstract: The use of learning-based methods for optimizing cellular radio access networks (RAN) has received increasing attention in recent years. This coincides with a rapid increase in the number of cell sites worldwide, driven largely by dramatic growth in cellular network traffic. Training and maintaining learned models that work well across a large number of cell sites has thus become a pertinent problem. This paper proposes a scalable framework for constructing a reinforcement learning policy bank that can perform RAN optimization across@inproceedings{rivkin2024cartier,
title={CARTIER: Cartographic lAnguage Reasoning Targeted at Instruction Execution for Robots},
author={Rivkin, Dmitriy and Kakodkar, Nikhil Rajiv and Hogan, Francois and Hamed Baghi, Bobak and Dudek, Gregory},
booktitle={Proc. International Conference on Robotics and Automation},
year={2024},
month={May}
} abstract: This work explores the capacity of large language models (LLMs) to address problems at the intersection of spatial planning and natural language interfaces for navigation. We focus on following complex instructions that are more akin to natural conversation than traditional explicit procedural directives typically seen in robotics. Unlike most prior work where navigation directives are provided as simple imperative commands (eg," go to the fridge"), we examine implicit directives obtained through conversational interactions. We leverage@inproceedings{chang2024imitation,
title={Imitation Learning from Observation through Optimal Transport},
author={Chang, Wei-Di and Fujimoto, Scott and Meger, David and Dudek, Gregory},
booktitle={Proceedings of the Reinforcement Learning Conference (RLC)},
year={2024},
note={to appear},
url={https://openreview.net/pdf?id=RI5frp6she}
} abstract: Imitation Learning from Observation (ILfO) is a setting in which a learner tries to imitate the behavior of an expert, using only observational data and without the direct guidance of demonstrated actions. In this paper, we re-examine the use of optimal transport for IL, in which a reward is generated based on the Wasserstein distance between the state trajectories of the learner and expert. We show that existing methods can be simplified to generate a reward function without requiring learned models or adversarial learning. Unlike@inproceedings{Jenkin2024,
author = {Michael Jenkin and Francois R. Hogan and Kaleem Siddiqi and Jean-François Tremblay and Bobak Baghi and Gregory Dudek},
title = {Interacting with a Visuotactile Countertop},
booktitle = {International Conference on Robotics, Computer Vision and Intelligent Systems},
pages = {361--374},
year = {2024},
publisher = {Springer Nature Switzerland},
address = {Cham}
} abstract: We present the See-Through-your-Skin Display (STS-d), a device that integrates visual and tactile sensing with a surface display to provide an interactive user experience. The STS-d expands the application of visuo-tactile optical sensors to Human-Robot Interaction (HRI) tasks and Human-Computer Interaction (HCI) tasks more generally. A key finding of this paper is that it is possible to display graphics on the reflective membrane of semi-transparent optical tactile sensors without interfering with their sensing capabilities@inproceedings{Xu2024,
author = {Yi Tian Xu and Di Wu and Michael Jenkin and Seowoo Jang and Xue Liu and Gregory Dudek},
title = {Optimizing Energy Saving for Wireless Networks Via Offline Decision Transformer},
booktitle = {Proc. ICC 2024 - IEEE International Conference on Communications},
year = {2024},
pages = {409--414},
doi = {10.1109/ICC51166.2024.10622786},
address = {Denver, CO, USA}
} abstract: With the global aim of reducing carbon emissions, energy saving for communication systems has gained tremendous attention. Efficient energy-saving solutions are not only required to accommodate the fast growth in communication demand but solutions are also challenged by the complex nature of the load dynamics. Recent reinforcement learning (RL)-based methods have shown promising performance for network optimization problems, such as base station energy saving. However, a major limitation of these methods is the requirement@inproceedings{lajoie2024peoplex,
author = {Pierre-Yves Lajoie and Bobak Hamed Baghi and Sachini Herath and Francois Hogan and Xue Liu and Gregory Dudek},
title = {PEOPLEx: PEdestrian Opportunistic Positioning LEveraging IMU, UWB, BLE and WiFi},
booktitle = {ICC 2024-IEEE International Conference on Communications},
pages = {3518--3523},
year = {2024},
publisher = {IEEE},
note = {ICC 2024 best paper award recipient}
} abstract: This paper advances the field of pedestrian localization by introducing a unifying framework for opportunistic positioning based on nonlinear factor graph optimization. While many existing approaches assume constant availability of one or multiple sensing signals, our methodology employs IMU-based pedestrian inertial navigation as the backbone for sensor fusion, opportunistically integrating Ultra-Wideband (UWB), Bluetooth Low Energy (BLE), and WiFi signals when they are available in the environment. The proposed PEOPLEx ...@inproceedings{Limoyo2024,
author = {O. Limoyo and J. Li and D. Rivkin and J. Kelly and G. Dudek},
title = {PhotoBot: Reference-Guided Interactive Photography via Natural Language},
booktitle = {Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
year = {2024},
url = {http://arxiv.org/abs/2401.11061},
note = {Also arXiv, Mar. 20, 2024}
} abstract: We introduce PhotoBot, a framework for automated photo acquisition based on an interplay between high-level human language guidance and a robot photographer. We propose to communicate photography suggestions to the user via a reference picture that is retrieved from a curated gallery. We exploit a visual language model (VLM) and an object detector to characterize reference pictures via textual descriptions and use a large language model (LLM) to retrieve relevant reference pictures based on a user's language query through text@inproceedings{al2024probabilistic,
title={Probabilistic Mobility Load Balancing for Multi-Band 5G and Beyond Networks},
author={Al Lahham, Saria and Wu, Di and Hossain, Ekram and Liu, Xue and Dudek, Gregory},
booktitle={2024 IEEE International Conference on Communications Workshops (ICC Workshops)},
pages={1673--1678},
year={2024},
organization={IEEE},
note={Also available: \url{http://arxiv.org/abs/2401.13792}}
} abstract: The ever-increasing demand for data services and the proliferation of user equipment (UE) have resulted in a significant rise in the volume of mobile traffic. Moreover, in multi-band networks, non-uniform traffic distribution among different operational bands can lead to congestion, which can adversely impact the user's quality of experience. Load balancing is a critical aspect of network optimization, where it ensures that the traffic is evenly distributed among different bands, avoiding congestion and ensuring bett er user experience@inproceedings{lotfi2024uncertainty,
title={Uncertainty-Aware Hybrid Paradigm of Nonlinear MPC and Model-Based RL for Offroad Navigation: Exploration of Transformers in the Predictive Model},
author={Lotfi, Faraz and Virji, Khalil and Faraji, Farnoosh and Berry, Lucas and Holliday, Andrew and Meger, David Paul and Dudek, Gregory},
booktitle={Proc. International Conference on Robotics and Automation},
year={2024},
month={May}
} abstract: In this paper, we investigate a hybrid scheme that combines nonlinear model predictive control (MPC) and model-based reinforcement learning (RL) for navigation planning of an autonomous model car across offroad, unstructured terrains without relying on predefined maps. Our innovative approach takes inspiration from BADGR, an LSTM-based network that primarily concentrates on environment modeling, but distinguishes itself by substituting LSTM modules with transformers to greatly elevate the performance of our model@inproceedings{limoyo2024working,
title={Working backwards: Learning to Place by Picking},
author={Limoyo, O. and Konar, A. and Ablett, T. and Kelly, J. and Hogan, F. and Dudek, G.},
booktitle={Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
year={2024}
} abstract: We present Learning to Place by Picking (LPP), a method capable of autonomously collecting demonstrations for a family of placing tasks in which objects must be manipulated to specific locations. With LPP, we approach the learning of robotic object placement policies by reversing the grasping process and exploiting the inherent symmetry of the pick and place problems. Specifically, we obtain placing demonstrations from a set of grasp sequences of objects that are initially located at their target placement locations. Our system@article{Ablett2024,
author = {T. Ablett and O. Limoyo and A. Sigal and A. Jilani and J. Kelly and K. Siddiqi and F. Hogan and G. Dudek},
title = {Multimodal and Force-Matched Imitation Learning with a See-Through Visuotactile Sensor},
journal = {arXiv preprint arXiv:2311.01248},
year = {2024},
url = {https://arxiv.org/abs/2311.01248}
} abstract: Multimodal and Force-Matched Imitation Learning with a See-Through Visuotactile Sensor@book{dudek2024computational,
title={Computational Principles of Mobile Robotics (3rd edition)},
author={Dudek, Gregory and Jenkin, Michael},
edition={3rd},
year={2024},
publisher={Cambridge University Press},
isbn={9780521692120},
pages={450}
} abstract: Computational Principles of Mobile Robotics 3rd edition@inproceedings{hu2023adateacher,
abstract = {To deal with notorious delays in communication systems, it is crucial to forecast key system characteristics, such as the communication load. Most existing studies aggregate data from multiple edge nodes for improving the forecasting accuracy. However, the bandwidth cost of such data aggregation could be unacceptably high from the perspective of system operators. To achieve both the high forecasting accuracy and bandwidth efficiency, this paper proposes an Adaptive Multi-Teacher Weighting in Teacher-Student Learning approach},
author = {Hu, Chengming and Wang, Ju and Wu, Di and Xin, Yan and Zhang, Charlie and Liu, Xue and Dudek, Gregory},
booktitle = {GLOBECOM 2023-2023 IEEE Global Communications Conference},
organization = {IEEE},
pages = {7399--7404},
pub_year = {2023},
title = {AdaTeacher: Adaptive Multi-Teacher Weighting for Communication Load Forecasting},
venue = {… 2023-2023 IEEE …}
} abstract: To deal with notorious delays in communication systems, it is crucial to forecast key system characteristics, such as the communication load. Most existing studies aggregate data from multiple edge nodes for improving the forecasting accuracy. However, the bandwidth cost of such data aggregation could be unacceptably high from the perspective of system operators. To achieve both the high forecasting accuracy and bandwidth efficiency, this paper proposes an Adaptive Multi-Teacher Weighting in Teacher-Student Learning approach@inproceedings{rivkin2023ansel,
abstract = {Our work examines the way in which large language models can be used for robotic planning and sampling in the context of automated photographic documentation. Specifically, we illustrate how to produce a photo-taking robot with an exceptional level of semantic awareness by leveraging recent advances in general purpose language (LM) and vision-language (VLM) models. Given a high-level description of an event we use an LM to generate a natural-language list of photo descriptions that one would expect a photographer},
author = {Rivkin, Dmitriy and Dudek, Gregory and Kakodkar, Nikhil and Meger, David and Limoyo, Oliver and Jenkin, Michael and Liu, Xue and Hogan, Francois},
booktitle = {2023 IEEE International Conference on Robotics and Automation (ICRA)},
organization = {IEEE},
pages = {8262--8268},
pub_year = {2023},
title = {Ansel photobot: A robot event photographer with semantic intelligence},
venue = {… on Robotics and …}
} abstract: Our work examines the way in which large language models can be used for robotic planning and sampling in the context of automated photographic documentation. Specifically, we illustrate how to produce a photo-taking robot with an exceptional level of semantic awareness by leveraging recent advances in general purpose language (LM) and vision-language (VLM) models. Given a high-level description of an event we use an LM to generate a natural-language list of photo descriptions that one would expect a photographer@inproceedings{zhao2023byzantine,
author = {Hanqing Zhao and Alexandre Pacheco and Volker Strobel and Andreagiovanni Reina and Xue Liu and Gregory Dudek and Marco Dorigo},
title = {A Generic Framework for ``Byzantine-tolerant Consensus Achievement in Robot Swarms''},
booktitle = {Proc. IEEE/RSJ International Conference on Robotics and Systems (IROS)},
year = {2023},
month = {Oct.},
pages = {8}
} abstract: Byzantine-tolerant Consensus Achievement in Robot Swarms@inproceedings{konar2023communication,
title={Communication Load Balancing via Efficient Inverse Reinforcement Learning},
author={Konar, Abhisek and Wu, Di and Xu, Yi Tian and Jang, Seowoo and Liu, Steve and Dudek, Gregory},
booktitle={IEEE International Conference on Communications (ICC)},
year={2023},
location={Rome, Italy}
} abstract: Communication load balancing aims to balance the load between different available resources, and thus improve the quality of service for network systems. After formulating the load balancing (LB) as a Markov decision process problem, reinforcement learning (RL) has recently proven effective in addressing the LB problem. To leverage the benefits of classical RL for load balancing, however, we need an explicit reward definition. Engineering this reward function is challenging, because it involves the need for expert knowledge and there@article{lotfi2024constrained,
abstract = {This paper explores leveraging large language models for map-free off-road navigation using generative AI, reducing the need for traditional data collection and annotation. We propose a method where a robot receives verbal instructions, converted to text through Whisper, and a large language model (LLM) model extracts landmarks, preferred terrains, and crucial adverbs translated into speed settings for constrained navigation. A language-driven semantic segmentation model generates text-based masks for identifying landmarks},
author = {Lotfi, Faraz and Faraji, Farnoosh and Kakodkar, Nikhil and Manderson, Travis and Meger, David and Dudek, Gregory},
journal = {arXiv preprint arXiv:2404.02294},
pub_year = {2024},
title = {Constrained Robotic Navigation on Preferred Terrains Using LLMs and Speech Instruction: Exploiting the Power of Adverbs},
venue = {arXiv preprint arXiv …}
} abstract: This paper explores leveraging large language models for map-free off-road navigation using generative AI, reducing the need for traditional data collection and annotation. We propose a method where a robot receives verbal instructions, converted to text through Whisper, and a large language model (LLM) model extracts landmarks, preferred terrains, and crucial adverbs translated into speed settings for constrained navigation. A language-driven semantic segmentation model generates text-based masks for identifying landmarks@article{wang2023eliminating,
abstract = {Due to their large bandwidth and impressive data speed, millimeter-wave (mmWave) radios are expected to play a key role in the 5G and beyond (eg, 6G) communication networks. Yet, to release mmWave's true power, the highly directional mmWave beams need to be aligned perfectly. Most existing beam alignment methods adopt an exhaustive or semi-exhaustive space scanning, which introduces up to seconds of delays. To eliminate the need for complex space scanning, this article presents an Ultra-wideband (UWB)-assisted mmWave},
author = {Wang, Ju and Chen, Xi and Liu, Xue and Dudek, Gregory},
journal = {ACM Transactions on Sensor Networks},
number = {4},
pages = {1--20},
pub_year = {2023},
publisher = {ACM New York, NY},
title = {Eliminating Space Scanning: Fast mmWave Beam Alignment with UWB Radios},
venue = {ACM Transactions on Sensor …},
volume = {19}
} abstract: Due to their large bandwidth and impressive data speed, millimeter-wave (mmWave) radios are expected to play a key role in the 5G and beyond (eg, 6G) communication networks. Yet, to release mmWave's true power, the highly directional mmWave beams need to be aligned perfectly. Most existing beam alignment methods adopt an exhaustive or semi-exhaustive space scanning, which introduces up to seconds of delays. To eliminate the need for complex space scanning, this article presents an Ultra-wideband (UWB)-assisted mmWave@inproceedings{wu2023energy,
abstract = {With the increasing use of data-intensive mobile applications and the number of mobile users, the demand for wireless data services has been increasing exponentially in recent years. In order to address this demand, a large number of new cellular base stations are being deployed around the world, leading to a significant increase in energy consumption and greenhouse gas emission. Consequently, energy consumption has emerged as a key concern in the fifth-generation (5G) network era and beyond. Reinforcement learning (RL)},
author = {Wu, Di and Xu, Yi Tian and Jenkin, Michael and Jang, Seowoo and Hossain, Ekram and Liu, Xue and Dudek, Gregory},
booktitle = {GLOBECOM 2023-2023 IEEE Global Communications Conference},
organization = {IEEE},
pages = {7019--7024},
pub_year = {2023},
title = {Energy Saving in Cellular Wireless Networks via Transfer Deep Reinforcement Learning},
venue = {… 2023-2023 IEEE …}
} abstract: With the increasing use of data-intensive mobile applications and the number of mobile users, the demand for wireless data services has been increasing exponentially in recent years. In order to address this demand, a large number of new cellular base stations are being deployed around the world, leading to a significant increase in energy consumption and greenhouse gas emission. Consequently, energy consumption has emerged as a key concern in the fifth-generation (5G) network era and beyond. Reinforcement learning (RL)@inproceedings{rezaei2023hypernetworks,
abstract = {In this paper, hypernetworks are trained to generate behaviors across a range of unseen task conditions, via a novel TD-based training objective and data from a set of near-optimal RL solutions for training tasks. This work relates to meta RL, contextual RL, and transfer learning, with a particular focus on zero-shot performance at test time, enabled by knowledge of the task parameters (also known as context). Our technical approach is based upon viewing each RL algorithm as a mapping from the MDP specifics to the near-optimal},
author = {Rezaei-Shoshtari, Sahand and Morissette, Charlotte and Hogan, Francois R and Dudek, Gregory and Meger, David},
booktitle = {Proceedings of the AAAI Conference on Artificial Intelligence},
number = {8},
pages = {9579--9587},
pub_year = {2023},
title = {Hypernetworks for zero-shot transfer in reinforcement learning},
venue = {Proceedings of the …},
volume = {37}
} abstract: In this paper, hypernetworks are trained to generate behaviors across a range of unseen task conditions, via a novel TD-based training objective and data from a set of near-optimal RL solutions for training tasks. This work relates to meta RL, contextual RL, and transfer learning, with a particular focus on zero-shot performance at test time, enabled by knowledge of the task parameters (also known as context). Our technical approach is based upon viewing each RL algorithm as a mapping from the MDP specifics to the near-optimal@inproceedings{wu2023learning,
abstract = {The association of mobile devices with network resources (eg, base stations, frequency bands/channels), known as load balancing, is critical to reduce communication traffic congestion and network performance. Reinforcement learning (RL) has shown to be effective for communication load balancing and achieves better performance than currently used rule-based methods, especially when the traffic load changes quickly. However, RL-based methods usually need to interact with the environment for a large number of time},
author = {Wu, Di and Xu, Yi Tian and Li, Jimmy and Jenkin, Michael and Hossain, Ekram and Jang, Seowoo and Xin, Yan and Zhang, Charlie and Liu, Xue and Dudek, Gregory},
booktitle = {GLOBECOM 2023-2023 IEEE Global Communications Conference},
organization = {IEEE},
pages = {2973--2978},
pub_year = {2023},
title = {Learning to Adapt: Communication Load Balancing via Adaptive Deep Reinforcement Learning},
venue = {… 2023-2023 IEEE …}
} abstract: The association of mobile devices with network resources (eg, base stations, frequency bands/channels), known as load balancing, is critical to reduce communication traffic congestion and network performance. Reinforcement learning (RL) has shown to be effective for communication load balancing and achieves better performance than currently used rule-based methods, especially when the traffic load changes quickly. However, RL-based methods usually need to interact with the environment for a large number of time@inproceedings{yuan2023mixed,
author = {Dun Yuan and Yujin Nam and Amal Feriani and Abhisek Konar and Di Wu and Seowoo Jang and Xue Liu and Greg Dudek},
title = {Mixed-Variable PSO with Fairness on Multi-Objective Field Data Replication in Wireless Networks},
booktitle = {Proc. IEEE International Conference on Communications (ICC)},
year = {2023}
} abstract: Digital twins have shown a great potential in supporting the development of wireless networks. They are virtual representations of 5G/6G systems enabling the design of machine learning and optimization-based techniques. Field data replication is one of the critical aspects of building a simulation-based twin, where the objective is to calibrate the simulation to match field performance measurements. Since wireless networks involve a variety of key performance indicators (KPIs), the replication process becomes a multi-objective@inproceedings{kang2023multi,
title={Multi-agent Attention Actor-Critic Algorithm for Load Balancing in Cellular Networks},
author={Kang, Jikun and Wu, Di and Wang, Ju and Hossain, Ekram and Liu, Xue and Dudek, Gregory},
booktitle={IEEE International Conference on Communications (ICC)},
year={2023},
address={Rome, Italy},
doi={10.48550/arXiv.2303.08003}
} abstract: In cellular networks, User Equipment (UE) handoff from one Base Station (BS) to another, giving rise to the load balancing problem among the BSs. To address this problem, BSs can work collaboratively to deliver a smooth migration (or handoff) and satisfy the UEs' service requirements. This paper formulates the load balancing problem as a Markov game and proposes a Robust Multi-agent Attention Actor-Critic (Robust-MA3C) algorithm that can facilitate collaboration among the BSs (ie, agents). In particular, to solve the Markov game@article{wu2023reinforcement,
abstract = {The amount of cellular communication network traffic has increased dramatically in recent years, and this increase has led to a demand for enhanced network performance. Communication load balancing aims to balance the load across available network resources and thus improve the quality of service for network users. Most existing load balancing algorithms are manually designed and tuned rule-based methods where near-optimality is almost impossible to achieve. Furthermore, rule-based methods are difficult to},
author = {Wu, Di and Li, Jimmy and Ferini, Amal and Xu, Yi Tian and Jenkin, Michael and Jang, Seowoo and Liu, Xue and Dudek, Gregory},
journal = {Frontiers in Computer Science},
pages = {1156064},
pub_year = {2023},
publisher = {Frontiers},
title = {Reinforcement learning for communication load balancing: approaches and challenges},
venue = {Frontiers in Computer …},
volume = {5}
} abstract: The amount of cellular communication network traffic has increased dramatically in recent years, and this increase has led to a demand for enhanced network performance. Communication load balancing aims to balance the load across available network resources and thus improve the quality of service for network users. Most existing load balancing algorithms are manually designed and tuned rule-based methods where near-optimality is almost impossible to achieve. Furthermore, rule-based methods are difficult to@inproceedings{lotfi2023robust,
abstract = {Target tracking is a classic problem in computer vision, with numerous applications in robotics. However, tracking targets underwater presents additional complications due to the six degrees of freedom nature of the problem and the challenging visual environment. In this paper, we address the problem of robotic underwater tracking of scuba divers by partitioning it into two parts: vision and control. We propose a new approach that exploits a highly-maneuverable underwater robot to perform experiments in open water, coupling sensing},
author = {Lotfi, Faraz and Virji, Khalil and Dudek, Gregory},
booktitle = {2023 20th Conference on Robots and Vision (CRV)},
organization = {IEEE},
pages = {233--240},
pub_year = {2023},
title = {Robust Scuba Diver Tracking and Recovery in Open Water Using YOLOv7, SORT, and Spiral Search},
venue = {2023 20th Conference on Robots and …}
} abstract: Target tracking is a classic problem in computer vision, with numerous applications in robotics. However, tracking targets underwater presents additional complications due to the six degrees of freedom nature of the problem and the challenging visual environment. In this paper, we address the problem of robotic underwater tracking of scuba divers by partitioning it into two parts: vision and control. We propose a new approach that exploits a highly-maneuverable underwater robot to perform experiments in open water, coupling sensing@inproceedings{kozlov2023self,
title={Self-Supervised Transformer Architecture for Change Detection in Radio Access Networks},
author={Kozlov, Igor and Rivkin, Dmitriy and Chang, Wei-Di and Wu, Di and Liu, Xue and Dudek, Gregory},
booktitle={Proc. IEEE International Conference on Communications (ICC)},
year={2023},
doi={10.48550/arXiv.2302.02025}
} abstract: Radio Access Networks (RANs) for telecommunications represent large agglomerations of interconnected hardware consisting of hundreds of thousands of transmitting devices (cells). Such networks undergo frequent and often heterogeneous changes caused by network operators, who are seeking to tune their system parameters for optimal performance. The effects of such changes are challenging to predict and will become even more so with the adoption of fifth-generation/sixth-generation (5G/6G) networks. Therefore, RAN monitoring is@inproceedings{zhao2023zero,
author = {Hanqing Zhao and Xue Liu and Gregory Dudek},
title = {Zero-shot Fault Detection for Manipulators through Bayesian Inverse Reinforcement Learning},
booktitle = {Proc. IEEE/RSJ International Conference on Robotics and Systems (IROS)},
year = {2023},
month = {Oct.},
pages = {8}
} abstract: We consider the detection of faults in robotic manipulators, with particular emphasis on faults that have not been observed or identified in advance, which naturally includes those that occur very infrequently. Recent studies indicate that the reward function obtained through Inverse Reinforcement Learning (IRL) can help detect anomalies caused by faults in a control system (ie fault detection). Current IRL methods for fault detection, however, either use a linear reward representation or require extensive sampling from the environment to@article{ablett2023push,
title={Push it to the Demonstrated Limit: Multimodal Visuotactile Imitation Learning with Force Matching},
author={Ablett, Trevor and Limoyo, Oliver and Sigal, Adam and Jilani, Affan and Kelly, Jonathan and Siddiqi, Kaleem and Hogan, Francois and Dudek, Gregory},
journal={arXiv preprint arXiv:2311.01248},
year={2023}
} abstract: Optical tactile sensors have emerged as an effective means to acquire dense contact information during robotic manipulation. A recently-introducedsee-through-your-skin'(STS) variant of this type of sensor has both visual and tactile modes, enabled by leveraging a semi-transparent surface and controllable lighting. In this work, we investigate the benefits of pairing visuotactile sensing with imitation learning for contact-rich manipulation tasks. First, we use tactile force measurements and a novel algorithm during kinesthetic teaching to yield@inproceedings{Kang2022,
author = {Jikun Kang and Ju Wang and Chengming Hu and Xue Liu and Gregory Dudek},
title = {A Generalized Load Balancing Policy With Multi-Teacher Reinforcement Learning},
booktitle = {Proceedings of GLOBECOM 2022-2022 IEEE Global Communications Conference},
year = {2022},
pages = {3096--3101},
publisher = {IEEE}
} abstract: Although reinforcement learning (RL) shows advantages in cellular network load balancing, it suffers from a low generalization ability, preventing it from real-world applications. Specifically, if network traffic pattern changes, the learned RL policy cannot adapt accordingly, resulting in system performance degradation. To address this issue, we propose a Multi-teacher MOdel BAsed Reinforcement Learning algorithm (MOBA), which leverages multi-teacher knowledge distillation theory to learn a generalized load balancing@inproceedings{hu2022accurate,
title={Accurate Communication Traffic Forecasting with Multi-Source Adaptive Feature Boosting},
author={Hu, Chengming and Wang, Ju and Wu, Di and Liu, Xue and Dudek, Gregory},
booktitle={Proc. GLOBECOM 2022-2022 IEEE Global Communications Conference},
pages={2316--2321},
year={2022},
organization={IEEE}
} abstract: Advanced communication network functions, such as resource allocation and dynamic spectrum management, heavily rely on the accurate forecasting of traffic. Data-driven solutions, eg, Neural Network (NN) based forecasting methods, have been proven to be effective only when sufficient data is available. However, Base Stations (BSs) have limited data in the real world, since big data for communication networks could be extremely expensive to collect, store, and migrate. Therefore, most existing traffic forecasting methods@inproceedings{wu2022active,
author = {Wu, D. and Jenkin, M. and Liu, X. and Dudek, G.},
title = {Active deep multi-task learning for forecasting short-term loads},
booktitle = {IEEE International Conference on Communications, ICC 2022},
year = {2022},
pages = {5523--5529},
doi = {10.1109/ICC45855.2022.9838341}
} abstract: With the increasing adoption of renewable energy generation and electric devices, electric load forecasting, especially short-term load forecasting (STLF), is becoming more and more important. The widespread adoption of smart meters makes it possible to utilize complex machine learning models for both aggregated load and single-home residential load forecasting. Similar homes in nearby locations are likely to have similar load consumption patterns and this similarity can be used to improve the overall forecasting performance@inproceedings{wu2022attentive,
title={Attentive Knowledge Transfer for Short-term Load Forecasting},
author={Wu, Di and Jenkin, Michael and Xu, Yi Tian and Liu, Xue and Dudek, Gregory},
booktitle={Proc GLOBECOM 2022-2022 IEEE Global Communications Conference},
pages={5285--5291},
year={2022},
organization={IEEE}
} abstract: The modern power system is transitioning towards increasing penetration of renewable energy generation and demand from different types of electrical appliances. With this transition, residential load forecasting, especially short-term load forecasting (STLF), is becoming more and more challenging and important. Accurate short-term load forecasting can help improve energy dispatching efficiency and, as a consequence, reduce overall power system operation cost. Most current load forecasting algorithms assume that there is a@inproceedings{zhao2022behaviour,
title={Behaviour Learning with Adaptive Motif Discovery and Interacting Multiple Model},
author={Zhao, Hanging and Manderson, Travis and Zhang, Hao and Liu, Xue and Dudek, Gregory},
booktitle={2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
pages={10788--10794},
year={2022},
organization={IEEE}
} abstract: We propose an approach that enables simultaneous interpretable learning of a high-level discrete behaviour and its low-level rhythmic sub-behaviour. We do this though a unified reward function, where a reward function that only describes low-level behaviour, with less impact on learning of other behaviours is recovered from few-shot motion demonstrations. To this end, we first extract local behaviour motifs from state-only human demonstrations and random driving samples using an adaptive motif discovery approach derived from the Matrix@inproceedings{li2022communication,
title={Communication Traffic Prediction with Continual Knowledge Distillation},
author={Li, Hang and Wang, Ju and Hu, Chengming and Chen, Xi and Liu, Xue and Jang, Seowoo and Dudek, Gregory},
booktitle={ICC 2022-IEEE International Conference on Communications},
pages={5481--5486},
year={2022},
publisher={IEEE}
} abstract: Accurate traffic volume estimation and prediction are essential for advanced communication network functions, such as automatic operations and predictive resource allocation. Although machine learning (ML)-based approaches achieve great success in accomplishing this goal, existing approaches suffer from two drawbacks that limit their real-world applications. First, the ML-based prediction models developed in the past might be obsolete now, since the communication traffic patterns and volumes keep changing in the real world@inproceedings{ma2022coordinated,
author = {Manyou Ma and Di Wu and Yi Tian Xu and Jimmy Li and Seowoo Jang and Xue Liu and Gregory Dudek},
title = {Coordinated Load Balancing in Mobile Edge Computing Network: a Multi-Agent DRL Approach},
booktitle = {ICC 2022-IEEE International Conference on Communications},
year = {2022},
pages = {619--624},
publisher = {IEEE}
} abstract: Mobile edge computing (MEC) networks have been recently adopted to accommodate the fast-growing number of mobile devices performing complicated tasks with limited hardware capability. Recently, edge nodes with communication, computation, and caching capacities are starting to be deployed in MEC networks. Due to the physical separation of these resources, efficient coordination and scheduling are important for efficient resource utilization and optimal network performance. In this paper, we study mobility load balancing@inproceedings{li2022data,
title={Data-Efficient Communication Traffic Prediction With Deep Transfer Learning},
author={Li, Hang and Wang, Ju and Chen, Xi and Liu, Xue and Dudek, Gregory},
booktitle={Proc ICC 2022-IEEE International Conference on Communications},
pages={3190--3195},
year={2022},
organization={IEEE}
} abstract: Prediction of future traffic load is a crucial task to support the automatic Operations, Administration, and Management (OAM) of communication networks. Existing Machine Learning (ML) models require big data to accomplish this task. However, large data sets are not always available, due to the limited storage capacity and the high storage cost at Base Stations (BSs). To solve the problem, we leverage the spatial-temporal correlation among different BSs, which allows other BSs' data to be used for the prediction of the target BS. One@inproceedings{hussien2022efficient,
title={Efficient Neural Data Compression for Machine Type Communications via Knowledge Distillation},
author={Hussien, Mostafa and Xu, Yi Tian and Wu, Di and Liu, Xue and Dudek, Gregory},
booktitle={GLOBECOM 2022-2022 IEEE Global Communications Conference},
pages={1169--1174},
year={2022},
publisher={IEEE}
} abstract: The anticipated huge number of devices and large traffic volumes impose new challenges on the communication system requirements and design. One of the main requirements of massive machine-type communication (mMTC) is to support network energy efficiency. Data compression is a widely adopted technique that enables higher energy efficiency, lower latency, and better bandwidth utilization. Unfortunately, the current compression techniques are mainly designed for human-type communications (HTC). Therefore, they consider the@article{chen2022fidora,
abstract = {Emerging Internet of Things (IoT) applications, such as cashier-less shopping, mobile ads targeting, and geo-based augmented reality (AR), are expected to bring us much more convenience and infotainment. To realize this amazing future, we need to feed these applications with user locations of (sub) meter-level resolution anytime and anywhere. Unfortunately, many widely used location sources are either unavailable indoor (eg, global positioning system) or coarse grained (eg, user check-ins). In order to provide ubiquitous},
author = {Chen, Xi and Li, Hang and Zhou, Chenyi and Liu, Xue and Wu, Di and Dudek, Gregory},
journal = {IEEE Internet of Things Journal},
number = {12},
pages = {9872--9888},
pub_year = {2022},
publisher = {IEEE},
title = {Fidora: Robust WiFi-based indoor localization via unsupervised domain adaptation},
venue = {IEEE Internet of Things …},
volume = {9}
} abstract: Emerging Internet of Things (IoT) applications, such as cashier-less shopping, mobile ads targeting, and geo-based augmented reality (AR), are expected to bring us much more convenience and infotainment. To realize this amazing future, we need to feed these applications with user locations of (sub) meter-level resolution anytime and anywhere. Unfortunately, many widely used location sources are either unavailable indoor (eg, global positioning system) or coarse grained (eg, user check-ins). In order to provide ubiquitous@article{hogan2022finger,
abstract = {This paper introduces and develops novel touch sensing technologies that enable robots to better sense and react to to intermittent contact interactions. We present Finger-STS, a robotic finger embodiment of the See-Through-your-Skin (STS) sensor that can capture 1) an “in the hand” visual perspective of an object that is being manipulated and 2) a high resolution tactile imprint of the contact geometry. We demonstrate the value of the sensor on a Bead Maze task. Here the multimodal feedback provided by the Finger-STS is leveraged},
author = {Hogan, Francois R and Tremblay, Jean-Fran{\c{c}}ois and Baghi, Bobak H and Jenkin, Michael and Siddiqi, Kaleem and Dudek, Gregory},
journal = {IEEE Robotics and Automation Letters},
number = {4},
pages = {10865--10872},
pub_year = {2022},
publisher = {IEEE},
title = {Finger-sts: Combined proximity and tactile sensing for robotic manipulation},
venue = {IEEE Robotics and …},
volume = {7}
} abstract: This paper introduces and develops novel touch sensing technologies that enable robots to better sense and react to to intermittent contact interactions. We present Finger-STS, a robotic finger embodiment of the See-Through-your-Skin (STS) sensor that can capture 1) an “in the hand” visual perspective of an object that is being manipulated and 2) a high resolution tactile imprint of the contact geometry. We demonstrate the value of the sensor on a Bead Maze task. Here the multimodal feedback provided by the Finger-STS is leveraged@article{chang2022ilflow,
title={Il-flow: Imitation learning from observation using normalizing flows},
author={Chang, Wei-Di and Gamboa Higuera, Juan Camilo and Fujimoto, Scott and Meger, David and Dudek, Gregory},
journal={arXiv preprint arXiv:2205.09251},
year={2022},
note={4th Robot Learning Workshop: Self-Supervised and Lifelong Learning, NeurIPS 2021}
} abstract: We present an algorithm for Inverse Reinforcement Learning (IRL) from expert state observations only. Our approach decouples reward modelling from policy learning, unlike state-of-the-art adversarial methods which require updating the reward model during policy search and are known to be unstable and difficult to optimize. Our method, IL-flOw, recovers the expert policy by modelling state-state transitions, by generating rewards using deep density estimators trained on the demonstration trajectories, avoiding the instability issues of@article{feriani2022multiobjective,
abstract = {Load balancing has become a key technique to handle the increasing traffic demand and improve the user experience. It evenly distributes the traffic across network resources by offloading users from overloaded base stations or channels to less crowded ones. Load balancing is a multi-objective optimization problem involving the automatic adjustment of several parameters to simultaneously maximize multiple network performance indicators. However, the existing methods mostly rely on single-objective approaches which lead to sub},
author = {Feriani, Amal and Wu, Di and Xu, Yi Tian and Li, Jimmy and Jang, Seowoo and Hossain, Ekram and Liu, Xue and Dudek, Gregory},
journal = {IEEE Journal on Selected Areas in Communications},
number = {9},
pages = {2614--2629},
pub_year = {2022},
publisher = {IEEE},
title = {Multiobjective load balancing for multiband downlink cellular networks: A meta-reinforcement learning approach},
venue = {IEEE Journal on …},
volume = {40}
} abstract: Load balancing has become a key technique to handle the increasing traffic demand and improve the user experience. It evenly distributes the traffic across network resources by offloading users from overloaded base stations or channels to less crowded ones. Load balancing is a multi-objective optimization problem involving the automatic adjustment of several parameters to simultaneously maximize multiple network performance indicators. However, the existing methods mostly rely on single-objective approaches which lead to sub@inproceedings{xu2022policy,
author = {Tian Xu and J. Li and D. Wu and M. Jenkin and S. Jang and X. Liu and G. Dudek},
title = {Policy Reuse for Communication Load Balancing in Unseen Traffic Scenarios},
booktitle = {Proc. IEEE International Conference on Communications (ICC)},
year = {2022},
pages = {619--624},
address = {Rome, Italy}
} abstract: With the continuous growth in communication network complexity and traffic volume, communication load balancing solutions are receiving increasing attention. Specifically, reinforcement learning (RL)-based methods have shown impressive performance compared with traditional rule-based methods. However, standard RL methods generally require an enormous amount of data to train, and generalize poorly to scenarios that are not encountered during training. We propose a policy reuse framework in which a policy@article{manjanna2022scalable,
abstract = {This paper presents a distributed scalable multi-robot planning algorithm for informed sampling of quasistatic spatials fields. We address the problem of efficient data collection using multiple autonomous vehicles and consider the effects of communication between multiple robots, acting independently, on the overall sampling performance of the team. We focus on the distributed sampling problem where the robots operate independent of their teammates, but have the ability to communicate their current state to other neighbors within},
author = {Manjanna, Sandeep and Hsieh, M Ani and Dudek, Greogory},
journal = {Autonomous Robots},
number = {7},
pages = {817--829},
pub_year = {2022},
publisher = {Springer},
title = {Scalable multirobot planning for informed spatial sampling},
venue = {Autonomous Robots},
volume = {46}
} abstract: This paper presents a distributed scalable multi-robot planning algorithm for informed sampling of quasistatic spatials fields. We address the problem of efficient data collection using multiple autonomous vehicles and consider the effects of communication between multiple robots, acting independently, on the overall sampling performance of the team. We focus on the distributed sampling problem where the robots operate independent of their teammates, but have the ability to communicate their current state to other neighbors within@inproceedings{Baghi2022,
author = {Baghi, Bobak H and Konar, Abhisek and Hogan, Francois and Jenkin, Michael and Dudek, Gregory},
title = {SESNO: Sample Efficient Social Navigation from Observation},
booktitle = {2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
year = {2022},
pages = {9164--9171},
publisher = {IEEE}
} abstract: In this paper, we present the Sample Efficient Social Navigation from Observation (SESNO) algorithm that efficiently learns socially-compliant navigation policies from observations of human trajectories. SESNO is an inverse reinforcement learning (IRL)-based algorithm that learns from human trajectory observations without knowledge of their actions. We improve the sample-efficiency over previous IRL-based methods by introducing a shared experience replay buffer that allows reuse of past trajectory experiences to estimate the policy and the@inproceedings{wu2022short,
title={Short-term load forecasting with deep boosting transfer regression},
author={Wu, Di and Xu, Yi Tian and Jenkin, Michael and Wang, Ju and Li, Hang and Liu, Xue and Dudek, Gregory},
booktitle={ICC 2022-IEEE International Conference on Communications},
pages={5530--5536},
year={2022},
organization={IEEE}
} abstract: With the increasing popularity of electric vehicles and the growing trend of working from home, electricity consumption in the residential sector is expected to continue to grow rapidly over the next few years. As a consequence, short-term residential load forecasting is becoming even more vital for the reliability and sustainability of the smart grid. Although deep learning models have shown impressive success in different areas including short-term electric load forecasting, such models require a large amount of training data. For many@inproceedings{Li2022,
author = {J. Li and D. Wu and Y.T. Xu and T. Li and J. Seowoo and X. Liu and G.L. Dudek},
title = {Traffic scenario clustering and load balancing with distilled reinforcement learning policies},
booktitle = {Proc. IEEE International Conference on Communications (ICC)},
year = {2022},
month = {Apr.},
date = {20},
} abstract: Due to the rapid increase in wireless communication traffic in recent years, load balancing is becoming increasingly important for ensuring the quality of service. However, variations in traffic patterns near different serving base stations make this task challenging. On one hand, crafting a single control policy that performs well across all base station sectors is often difficult. On the other hand, maintaining separate controllers for every sector introduces overhead, and leads to redundancy if some of the sectors experience similar traffic patterns@inproceedings{Hansen2022,
author = {Hansen, J. and Hogan, F. and Rivkin, D. and Meger, D. and Jenkin, M. and Dudek, G.},
title = {Visuotactile-RL: Learning multimodal manipulation policies with deep reinforcement learning},
booktitle = {2022 International Conference on Robotics and Automation (ICRA)},
year = {2022},
pages = {8298--8304},
doi = {10.1109/ICRA46639.2022.9812019}
} abstract: Manipulating objects with dexterity requires timely feedback that simultaneously leverages the senses of vision and touch. In this paper, we focus on the problem setting where both visual and tactile sensors provide pixel-level feedback for Visuotactile reinforcement learning agents. We investigate the challenges associated with multimodal learning and propose several improvements to existing RL methods; including tactile gating, tactile data augmentation, and visual degradation. When compared with visual-only and tactile-only@inproceedings{Kang2021,
author = {J. Kang and X. Chen and D. Wu and Y.T. Xu and X. Liu and G. Dudek and T. Lee and I. Park},
title = {Hierarchical policy learning for hybrid communication load balancing},
booktitle = {Proc. IEEE International Conference on Communications (ICC)},
year = {2021},
pages = {6},
doi = {10.1109/ICC42927.2021.9500379}
} abstract: Due to the uneven demographic distribution and people's daily activities, communication systems usually experience highly imbalanced load across different cells. This imbalance leads to unsatisfied users in the congested cells and under-utilized resources in the less-loaded cells. To deal with this issue, existing work migrates the load from heavily loaded cells to lightly loaded cells, by either handing over active mode User Equipment (UEs) to other serving cells, or re-selecting the camping cells for idle mode UEs. In this paper, we@inproceedings{rivkin2021learning,
title={Learning Assisted Identification of Scenarios Where Network Optimization Algorithms Under-Perform},
author={Rivkin, Dmitriy and Meger, David and Wu, Di and Chen, Xi and Liu, ue and Dudek, Gregory},
booktitle={Proc. IEEE Global Communications Conference (Globecom 2021)},
pages={6},
year={2021},
address={Madrid, Spain},
month={Dec}
} abstract: We present a generative adversarial method that uses deep learning to identify network load traffic conditions in which network optimization algorithms under-perform other known algorithms: the Deep Convolutional Failure Generator (DCFG). The spatial distribution of network load presents challenges for network operators for tasks such as load balancing, in which a network optimizer attempts to maintain high quality communication while at the same time abiding capacity constraints. Testing a network optimizer for all possible load@article{konar2021learning,
abstract = {One of the main challenges of operating mobile robots in social environments is the safe and fluid navigation therein, specifically the ability to share a space with other human inhabitants by complying with the explicit and implicit rules that we humans follow during navigation. While these rules come naturally to us, they resist simple and explicit definitions. In this letter, we present a learning-based solution to address the question of socially compliant navigation, which is to navigate while maintaining adherence to the navigational},
author = {Konar, Abhisek and Baghi, Bobak H and Dudek, Gregory},
journal = {IEEE Robotics and Automation Letters},
number = {2},
pages = {651--658},
pub_year = {2021},
publisher = {IEEE},
title = {Learning goal conditioned socially compliant navigation from demonstration using risk-based features},
venue = {IEEE Robotics and Automation …},
volume = {6}
} abstract: One of the main challenges of operating mobile robots in social environments is the safe and fluid navigation therein, specifically the ability to share a space with other human inhabitants by complying with the explicit and implicit rules that we humans follow during navigation. While these rules come naturally to us, they resist simple and explicit definitions. In this letter, we present a learning-based solution to address the question of socially compliant navigation, which is to navigate while maintaining adherence to the navigational@inproceedings{Rezaei-Shoshtari2021,
author = {S. Rezaei-Shoshtari and F. R. Hogan and M. Jenkin and D. Meger and G. Dudek},
title = {Learning Intuitive Physics with Multimodal Generative Models},
booktitle = {Proceedings of the AAAI Conference on Artificial Intelligence},
year = {2021},
month = {Feb},
volume = {2101},
note = {preprint as doi:arXiv:2101.04454},
url = {http://adsabs.harvard.edu/abs/2021arXiv210104454R}
} abstract: Predicting the future interaction of objects when they come into contact with their environment is key for autonomous agents to take intelligent and anticipatory actions. This paper presents a perception framework that fuses visual and tactile feedback to make predictions about the expected motion of objects in dynamic scenes. Visual information captures object properties such as 3D shape and location, while tactile information provides critical cues about interaction forces and resulting object motion when it makes contact with@inproceedings{wu2021load,
author = {Di Wu and Jikun Kang and Yi Tian Xu and Hang Li and Jimmy Li and Xi Chen and Dmitriy Rivkin and Michael Jenkin and Taeseop Lee and Intaik Park and Xue Liu and Gregory Dudek},
title = {Load Balancing for Communication Networks via Data-Efficient Deep Reinforcement Learning},
booktitle = {Proc. IEEE Global Communications Conference (Globecom 2021)},
year = {2021},
address = {Madrid, Spain},
month = {Dec.},
pages = {6}
} abstract: Within a cellular network, load balancing between different cells is of critical importance to network performance and quality of service. Most existing load balancing algorithms are manually designed and tuned rule-based methods where near-optimality is almost impossible to achieve. These rule-based meth-ods are difficult to adapt quickly to traffic changes in real-world environments. Given the success of Reinforcement Learning (RL) algorithms in many application domains, there have been a number of efforts to tackle load@inproceedings{Tremblay2021,
author = {J.-F. Tremblay and T. Manderson and A. Noca and G. Dudek and D. Meger},
title = {Multimodal dynamics modeling for off-road autonomous vehicles},
booktitle = {2021 IEEE International Conference on Robotics and Automation (ICRA)},
year = {2021},
pages = {1796--1802},
doi = {10.1109/ICRA48506.2021.9561910}
} abstract: Dynamics modeling in outdoor and unstructured environments is difficult because different elements in the environment interact with the robot in ways that can be hard to predict. Leveraging multiple sensors to perceive maximal information about the robot's environment is thus crucial when building a model to perform predictions about the robot's dynamics with the goal of doing motion planning. We design a model capable of long-horizon motion predictions, leveraging vision, lidar and proprioception, which is robust to arbitrarily missing@inproceedings{chen2021one,
author = {Xi Chen and Ju Wang and Hang Li and Yi Tian Xu and Di Wu and Xue Liu and Gregory Dudek and Taeseop Lee and Intaik Park},
title = {One for All: Traffic Prediction at Heterogeneous 5G Edge with Data-Efficient Transfer Learning},
booktitle = {Proc. IEEE Global Communications Conference (Globecom 2021)},
year = {2021},
address = {Madrid, Spain},
month = {Dec.},
pages = {6},
note = {Best paper award}
} abstract: By placing the computing, storage and networking resources close to the end users, distributed edge computing greatly benefits the performance of 5G communication systems. However, as a tradeoff, resources on the edge are usually limited and imbalanced among the heterogeneous edge nodes. To overcome this drawback, this paper proposes a Transfer Learning based Prediction (TLP) framework that allows the edge nodes to share their resources and data in an efficient manner. In particular, the TLP framework focuses on the@inproceedings{Girdhar2021,
author = {Y. Girdhar and D. Rivkin and D. Wu and M. Jenkin and X. Liu and G. Dudek},
title = {Optimizing Cellular Networks via Continuously Moving Base Stations on Road Networks},
booktitle = {2021 IEEE International Conference on Robotics and Automation (ICRA)},
year = {2021},
pages = {4020--4025},
doi = {10/gntw2n}
} abstract: Although existing cellular network base stations are typically immobile, the recent development of small form factor base stations and self driving cars has enabled the possibility of deploying a team of continuously moving base stations that can reorganize the network infrastructure to adapt to changing network traffic usage patterns. Given such a system of mobile base stations (MBSes) that can freely move on the road, how should their path be planned in an effort to optimize the experience of the users? This paper addresses@article{holliday2021scale,
abstract = {This work presents Object Landmarks, a new type of visual feature designed for visual localization over major changes in distance and scale. An Object Landmark consists of a bounding box bb defining an object, a descriptor qq of that object produced by a Convolutional Neural Network, and a set of classical point features within b b. We evaluate Object Landmarks on visual odometry and place-recognition tasks, and compare them against several modern approaches. We find that Object Landmarks enable superior},
author = {Holliday, Andrew and Dudek, Gregory},
journal = {Autonomous Robots},
number = {3},
pages = {407--420},
pub_year = {2021},
publisher = {Springer},
title = {Scale-invariant localization using quasi-semantic object landmarks},
venue = {Autonomous Robots},
volume = {45}
} abstract: This work presents Object Landmarks, a new type of visual feature designed for visual localization over major changes in distance and scale. An Object Landmark consists of a bounding box bb defining an object, a descriptor qq of that object produced by a Convolutional Neural Network, and a set of classical point features within b b. We evaluate Object Landmarks on visual odometry and place-recognition tasks, and compare them against several modern approaches. We find that Object Landmarks enable superior@inproceedings{hogan2021seeing,
title={Seeing Through your Skin: Recognizing Objects with a Novel Visuotactile Sensor},
author={Hogan, F.R. and Jenkin, M. and Rezaei-Shoshtari, S. and Girdhar, Y. and Meger, D. and Dudek, G.},
booktitle={Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision},
pages={1218--1227},
year={2021}
} abstract: We introduce a new class of vision-based sensor and associated algorithmic processes that combine visual imaging with high-resolution tactile sending, all in a uniform hardware and computational architecture. We demonstrate the sensor's efficacy for both multi-modal object recognition and metrology. Object recognition is typically formulated as an unimodal task, but by combining two sensor modalities we show that we can achieve several significant performance improvements. This sensor, named the See-Through-your-Skin sensor (STS)@inproceedings{konar2021toy,
author = {Konar, A. and Baghi, B.H. and Hogan, F.R. and Dudek, G.},
title = {Toy Story1 Method for Multi-Policy Social Navigation},
booktitle = {RSS Workshop on Social Robot Navigation},
pages = {5},
year = {2021},
month = {Jul}
} abstract: In this work, we present a method for effective social navigation by selecting navigation policies based on local social context. Learning robotic navigation policies that are consistent with inferred social norms is challenging as these norms are often subjective, culturally dependent, task specific, and context sensitive. While learning-based approaches to social navigation have shown success, they must also be able to compete with established classical algorithms that confer theoretical and practical advantages in simpler@inproceedings{wang2021uwb,
author = {Wang, J. and Chen, X. and Liu, X. and Dudek, G.},
title = {UWB-Assisted Fast mmWave Beam Alignment},
booktitle = {ICC 2021 - IEEE International Conference on Communications},
year = {2021},
pages = {1--6},
doi = {10.1109/ICC42927.2021.9500352},
month = {Jun.}
} abstract: Due to their large bandwidth and impressive data speed, millimeter-wave (mmWave) radios are expected to play a key role in the 5G and beyond (eg, 6G) communication networks. Yet, to release mmWave's true power, the highly directional mmWave beams need to be aligned perfectly. Most existing beam alignment methods adopt an exhaustive or semi-exhaustive space scanning, which introduces up to seconds of delays. To eliminate the need of a complex space scanning, this paper presents an Ultra-wideband (UWB)-assisted mmWave@article{Manjanna2021,
author = {S. Manjanna and M. A. Hsieh and G. Dudek},
title = {Scalable Multi-Robot System for Non-myopic Spatial Sampling},
journal = {arXiv:2105.10018 [cs]},
year = {2021},
month = {Oct.},
note = {Accessed: Dec. 21, 2021},
url = {http://arxiv.org/abs/2105.10018}
} abstract: Scalable Multi-Robot System for Non-myopic Spatial Sampling@inproceedings{Friedman2020,
author = {Friedman, N. and Goedicke, D. and Zhang, V. and Rivkin, D. and Jenkin, M. and Degutyte, Z. and Astell, A. and Liu, X. and Dudek, G.},
title = {Capturing attention with wind},
booktitle = {Workshop on Approaches to Advance Physical Human-Robot Interaction (AVHC)},
year = {2020},
month = {May},
pages = {2},
url = {https://vgrserver.eecs.yorku.ca/~jenkin/papers/2020/2020ICRAWorkshop.pdf}
} abstract: Having a robot interact with people in a shared environment is complex. Both running into humans and loud audio warnings are inappropriate. Visual signalling may be appropriate but is only effective if the humans are looking at/attending to the robot vehicle. Are there effective and socially acceptable mechanisms that a robot can exploit to capture the attention of humans in a shared environment? Here we explore the potential of using controlled blasts of wind (haptic air) to capture attention in a socially acceptable manner.@inproceedings{gamboa2020collaborative,
author = {Juan Camilo Gamboa Higuera and Travis Manderson and Karim Koreitem and Wei-Di Chang and Florian Shkurti and David Meger and Gregory Dudek},
title = {Collaborative Human-Robot Exploration for Marine Environments},
booktitle = {RSS '20 Workshop on Assistive and Collaborative Robotics: Decoding Intent},
year = {2020}
} abstract: Collaborative Human-Robot Exploration for Marine Environments@inproceedings{joshi2020deepurl,
title={DeepURL: Deep Pose Estimation Framework for Underwater Relative Localization},
author={Joshi, Bharat and Modasshir, Md and Manderson, Travis and Damron, Hunter and Xanthidis, Marios and Quattrini Li, Alberto and Rekleitis, Ioannis and Dudek, Gregory},
booktitle={Proc. IEEE/RSJ International Conference on Robotics and Systems (IROS)},
year={2020}
} abstract: In this paper, we propose a real-time deep learning approach for determining the 6D relative pose of Autonomous Underwater Vehicles (AUV) from a single image. A team of autonomous robots localizing themselves in a communication-constrained underwater environment is essential for many applications such as underwater exploration, mapping, multi-robot convoying, and other multi-robot tasks. Due to the profound difficulty of collecting ground truth images with accurate 6D poses underwater, this work utilizes rendered images@inproceedings{Cheng2020,
author = {Ran Cheng and Christopher Agia and David Meger and Gregory Dudek},
title = {Depth Prediction for Monocular Direct Visual Odometry},
booktitle = {2020 17th Conference on Computer and Robot Vision (CRV)},
pages = {70--77},
year = {2020},
publisher = {IEEE Computer Society}
} abstract: Depth prediction from monocular images with deep CNNs is a topic of increasing interest to the community. Advances have lead to models capable of predicting disparity maps with consistent scale, which are an acceptable prior for gradient-based direct methods. With this in consideration, we exploit depth prediction as a candidate prior for the coarse initialization, tracking, and marginalization steps of the direct visual odometry system, enabling the second-order optimizer to converge faster into a precise global minimum. In addition, the@inproceedings{di2020dynamic,
title={Dynamic Planning of Redundant Robots within a Set-Based Task-Priority Inverse Kinematics Framework},
author={Di Vito, Daniele and Mathieux Bergeron, David Meger, Gregory Dudek, and Gianluca Antonelli},
booktitle={2020 IEEE Conference on Control Technology and Applications (CCTA)},
pages={549--554},
year={2020},
organization={IEEE}
} abstract: This work presents the dynamic planning of redundant robots by merging a global and local planner. The global planner is implemented as a sampling-based algorithm which works in the reduced-dimensionality of the robot workspace applying the Cartesian constraints only. The output trajectory is then checked within a framework of set-based task priority inverse kinematics verifying the fulfillment of the other task constraints. The inverse kinematics framework is used also in real-time as local motion control to ensure a reactive behaviour to@inproceedings{Chen2020FiDo,
author = {Xi Chen and Hang Li and Chenyi Zhou and Xue Liu and Di Wu and Gregory Dudek},
title = {FiDo: Ubiquitous Fine-Grained WiFi-Based Localization for Unlabelled Users via Domain Adaptation},
booktitle = {Proceedings of The Web Conference 2020 (WWW '20)},
pages = {23--33},
year = {2020},
location = {Taipei, Taiwan},
publisher = {Association for Computing Machinery},
doi = {10.1145/3366423.3380091}
} abstract: To fully support the emerging location-aware applications, location information with meter-level resolution (or even higher) is required anytime and anywhere. Unfortunately, most of the current location sources (eg, GPS and check-in data) either are unavailable indoor or provide only house-level resolutions. To fill the gap, this paper utilizes the ubiquitous WiFi signals to establish a (sub) meter-level localization system, which employs WiFi propagation characteristics as location fingerprints. However, an unsolved issue of these WiFi@inproceedings{manderson2020learning,
title={Learning to Drive Off-Road on Smooth Terrains in Unstructured Environments Using an Onboard Camera and Sparse Aerial Images},
author={Manderson, Travis and Wapnick, Stefan and Meger, Dave and Dudek, Gregory},
booktitle={Proceedings of the 2020 IEEE International Conference on Robotics and Automation},
year={2020}
} abstract: We present a method for learning to drive on smooth terrain while simultaneously avoiding collisions in challenging off-road and unstructured outdoor environments using only visual inputs. Our approach applies a hybrid model-based and model-free reinforcement learning method that is entirely self-supervised in labeling terrain roughness and collisions using on-board sensors. Notably, we provide both first-person and overhead aerial image inputs to our model. We nd that the fusion of these complementary inputs improves planning foresight@inproceedings{manderson2020learning,
title={Learning to Drive Off-Road on Smooth Terrains in Unstructured Environments Using an Onboard Camera and Sparse Aerial Images},
author={Manderson, Travis and Wapnick, Stefan and Meger, Dave and Dudek, Gregory},
booktitle={Proceedings of the 2020 IEEE International Conference on Robotics and Automation},
year={2020}
} abstract: We present a method for learning to drive on smooth terrain while simultaneously avoiding collisions in challenging off-road and unstructured outdoor environments using only visual inputs. Our approach applies a hybrid model-based and model-free reinforcement learning method that is entirely self-supervised in labeling terrain roughness and collisions using on-board sensors. Notably, we provide both first-person and overhead aerial image inputs to our model. We nd that the fusion of these complementary inputs improves planning foresight@inproceedings{koreitem2020one,
title={One-Shot Informed Robotic Visual Search in the Wild},
author={Koreitem, Karim and Shkurti, Florian and Manderson, Travis and Chang, Wei-Di and Gamboa Higuera, Juan Camilo and Dudek, Gregory},
booktitle={Proc. IEEE/RSJ International Conference on Robotics and Systems (IROS 2020)},
year={2020}
} abstract: We consider the task of underwater robot navigation for the purpose of collecting scientifically relevant video data for environmental monitoring. The majority of field robots that currently perform monitoring tasks in unstructured natural environments navigate via path-tracking a pre-specified sequence of waypoints. Although this navigation method is often necessary, it is limiting because the robot does not have a model of what the scientist deems to be relevant visual observations. Thus, the robot can neither visually search for@inproceedings{Friedman2020,
author = {Friedman, N. and Goedicke, D. and Zhang, V. and Rivkin, D. and Jenkin, M. and Degutyte, Z. and Astell, A. and Liu, X. and Dudek, G.},
title = {Out of my way! Exploring Different Modalities for Robots to Ask People to Move Out of the Way},
booktitle = {Workshop on Active Vision and Perception in Human(-Robot) Collaboration. Held in conjunction with the 29th IEEE Int. Conf. on Robot and Human Interactive Communication},
pages = {9},
year = {2020},
note = {Best paper award},
url = {https://vgrserver.eecs.yorku.ca/~jenkin/papers/2020/AVHRC2020-out-of-my-way.pdf}
} abstract: To navigate politely through social spaces, a mobile robot needs to communicate successfully with human bystanders. What is the best way for a robot to attract attention in a socially acceptable manner to communicate its intent to others in a shared space? Through a series of in-the-wild experiments, we measured the social appropriateness and effectiveness of different modalities for robots to communicate to people their intended movement, using combinations of visual text, audio and haptic cues. Using multiple@inproceedings{holliday2020pre,
title={Pre-Trained CNNs as Visual Feature Extractors: A Broad Evaluation},
author={Holliday, Andrew and Dudek, Gregory},
booktitle={2020 17th Conference on Computer and Robot Vision (CRV)},
pages={78--84},
year={2020},
organization={IEEE}
} abstract: In this work, we perform a wide-ranging evaluation of Convolutional Neural Networks (CNNs) as feature extractors for matching visual features under large changes in appearance, perspective, and visual scale. Our evaluation covers 82 different layers from twelve different CNN architectures belonging to four families: AlexNets, VGG Nets, ResNets, and DenseNets. To our knowledge, this is the most comprehensive analysis of its kind in the literature. We find that the intermediate layers of DenseNets serve as the best feature@inproceedings{Xu2020,
author = {Y.T. Xu and X. Chen and X. Liu and D. Meger and G. Dudek},
title = {PresSense: Passive Respiration Sensing via Ambient WiFi Signals in Noisy Environments},
booktitle = {2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
year = {2020},
month = {Oct},
pages = {4032--4039},
doi = {10/gntw35}
} abstract: Passive sensing with ambient WiFi signals is a promising technique that will enable new types of human-robot interactions while preserving users' privacy. Here, we present PresSense, a system for human respiration sensing in noisy environments. Unlike existing WiFi-based respiration sensors, we employ a human presence detector, improving the robustness in scenarios where no human is present in an Area Of Interest (AOI). We also integrate our novel feature, Peak Distance Histogram (PDH), with other classic WiFi features@inproceedings{hogan2020seeing,
title={Seeing Through Your Skin: A Novel Visuo-Tactile Sensor for Robotic Manipulation},
author={Hogan, F. R. and Rezaei-Shoshtari, S. and Jenkin, M. and Girdhar, Y. and Meger, D. and Dudek, G.},
booktitle={Visual Learning and Reasoning for Robotic Manipulation (Workshop of RSS 2020)},
year={2020},
address={Corvallis, Oregon, USA}
} abstract: Seeing Through Your Skin: A Novel Visuo-Tactile Sensor for Robotic Manipulation@inproceedings{manderson2020self,
title={Self-Supervised, Goal-Conditioned Policies for Navigation in Unstructured Environments},
author={Manderson, Travis and Gamboa Higuera, Juan Camilo and Wapnick, Stefan and Tremblay, Jean-Francois and Shkurti, Florian and Meger, David and Dudek, Gregory},
booktitle={Robotics Science and Systems (RSS) Workshop on Self-Supervised Robot Learning},
year={2020},
note={Best Paper Award}
} abstract: Self-Supervised, Goal-Conditioned Policies for Navigation in Unstructured Environments@inproceedings{jimmy2020view,
title={View-Invariant Loop Closure with Oriented Semantic Landmarks},
author={Jimmy, Karim Koreitem and Meger, David and Dudek, Gregory},
booktitle={2020 IEEE International Conference on Robotics and Automation (ICRA)},
pages={7943--7949},
year={2020},
organization={IEEE}
} abstract: Recent work on semantic simultaneous localization and mapping (SLAM) have shown the utility of natural objects as landmarks for improving localization accuracy and robustness. In this paper we present a monocular semantic SLAM system that uses object identity and inter-object geometry for view-invariant loop detection and drift correction. Our system's ability to recognize an area of the scene even under large changes in viewing direction allows it to surpass the mapping accuracy of ORB-SLAM, which uses only local appearance-based@inproceedings{manderson2020vision,
title={Vision-Based Goal-Conditioned Policies for Underwater Navigation in the Presence of Obstacles},
author={Manderson, Travis and Gamboa Higuera, Juan Camilo and Wapnick, Stefan and Tremblay, Jean-Francois and Shkurti, Florian and Meger, David and Dudek, Gregory},
booktitle={Proceeding of Robotics Science and Systems},
volume={16},
year={2020}
} abstract: We present Nav2Goal, a data-efficient and end-to-end learning method for goal-conditioned visual navigation. Our technique is used to train a navigation policy that enables a robot to navigate close to sparse geographic waypoints provided by a user without any prior map, all while avoiding obstacles and choosing paths that cover user-informed regions of interest. Our approach is based on recent advances in conditional imitation learning. General-purpose, safe and informative actions are demonstrated by a human expert. The learned@inproceedings{dudek2019environet,
title={EnviroNet: ImageNet Analog for Environment and Global AI Challenge},
author={Dudek, Gregory and Joppa, Lucas and Lakshmanan, Valliappa and Kumar, Vipin and Mukkavilli, Surya Karthik},
booktitle={99th American Meteorological Society Annual Meeting},
year={2019},
organization={AMS}
} abstract: 99th American Meteorological Society Annual Meeting@inproceedings{abeysirigoonawardena2019generating,
title={Generating Adversarial Driving Scenarios in High-Fidelity Simulators},
author={Abeysirigoonawardena, Yasasa and Shkurti, Florian and Dudek, Gregory},
booktitle={2019 International Conference on Robotics and Automation (ICRA)},
pages={8271--8277},
year={2019},
organization={IEEE}
} abstract: In recent years self-driving vehicles have become more commonplace on public roads, with the promise of bringing safety and efficiency to modern transportation systems. Increasing the reliability of these vehicles on the road requires an extensive suite of software tests, ideally performed on high-fidelity simulators, where multiple vehicles and pedestrians interact with the self-driving vehicle. It is therefore of critical importance to ensure that self-driving software is assessed against a wide range of challenging simulated driving@inproceedings{manderson2019heterogeneous,
title={Heterogeneous Robot Teams for Informative Sampling},
author={Manderson, Travis and Majanna, Sandeep and Dudek, Gregory},
booktitle={2019 Workshop on Informative Path Planning and Adaptive Sampling at Robotics Science and Systems},
year={2019},
url={https://arxiv.org/abs/1906.07208}
} abstract: In this paper we present a cooperative multi-robot strategy to adaptively explore and sample environments that are unfavorable for humans. We propose a methodology for a team of heterogeneous robots to collaborate on information based planning for applications like sampling thermal imagery in a wildfire affected site to assist with detecting spot fires and areas of residual fires, fire mapping and monitoring fire progression or applications in marine domain for coral reef monitoring and survey. We use Gabor filter based texture@inproceedings{wong2019investigating,
title={Investigating Trust Factors in Human-Robot Shared Control: Implicit Gender Bias Around Robot Voice},
author={Wong, Alex and Xu, Anqi and Dudek, Gregory},
booktitle={2019 16th Conference on Computer and Robot Vision (CRV)},
pages={195--200},
year={2019},
organization={IEEE}
} abstract: This paper explores the impact of warnings, audio feedback, and gender on human-robot trust in the context of autonomous driving and specifically shared robot control. We use pre-existing methods for the estimation and assessment of human-robot trust where trust was found to vary as a function of the quality of behavior of an autonomous driving controller. We extend these models and empirical methods to examine the impact of audio cues on trust, specifically studying the impacts of gender-specific audio cues on the elicitation of trust. Our@inproceedings{kumar2019physics,
title={Physics Guided ML: Emerging AI Opportunities for Weather and Climate},
author={Kumar, Vipin and Lakshmanan, Valliappa and Joppa, Lucas and Dudek, Gregory and Mukkavilli, Surya Karthik and McGovern, Amy},
booktitle={99th American Meteorological Society Annual Meeting},
year={2019},
organization={AMS},
note={Side panel: ``Side Panel Towards Planetary Intelligence: On the Synergistic Future of AI, Weather and Climate''}
} abstract: 99th American Meteorological Society Annual Meeting@inproceedings{manjanna2019policy,
author = {Sandeep Manjanna and Herke van Hoof and Gregory Dudek},
title = {Policy Search with Non-uniform State Representations for Environmental Sampling},
booktitle = {International Conference on Machine Learning, workshop on Climate Change, in association with NeurIPS 2019},
year = {2019},
url = {https://s3.us-east-1.amazonaws.com/climate-change-ai/papers/icml2019/1/paper.pdf}
} abstract: Surveying fragile ecosystems like coral reefs is important to monitor the effects of climate change. We present an adaptive sampling technique that generates efficient trajectories covering hotspots in the region of interest at a high rate. A key feature of our sampling algorithm is the ability to generate action plans for any new hotspot distribution using the parameters learned on other similar looking distributions.@inproceedings{li2019semantic,
title={Semantic Mapping for View-Invariant Relocalization},
author={Li, Jimmy and Meger, David and Dudek, Gregory},
booktitle={2019 International Conference on Robotics and Automation (ICRA)},
pages={7108--7115},
year={2019},
organization={IEEE}
} abstract: We propose a system for visual simultaneous localization and mapping (SLAM) that combines traditional local appearance-based features with semantically meaningful object landmarks to achieve both accurate local tracking and highly view-invariant object-driven relocalization. Our mapping process uses a sampling-based approach to efficiently infer the 3D pose of object landmarks from 2D bounding box object detections. These 3D landmarks then serve as a view-invariant representation which we leverage to achieve camera@inproceedings{koreitem2019underwater,
author = {Karim Koreitem and Jimmy Li and Ian Karp and Travis Manderson and Gregory Dudek},
title = {Underwater Communication Using Full-Body Gestures and Optimal Variable-Length Prefix Codes},
booktitle = {Proceedings of the 2019 IEEE International Conference on Robotics and Automation},
year = {2019}
} abstract: In this paper we consider inter-robot communication in the context of joint activities. In particular, we focus on convoying and passive communication for radio-denied environments by using whole-body gestures to provide cues regarding future actions. We develop a communication protocol whereby information described by codewords is transmitted by a series of actions executed by a swimming robot. These action sequences are chosen to optimize robustness and transmission duration given the observability, natural@inproceedings{Manjanna2018,
author = {Sandeep Manjanna and Alberto Quattrini Li and Ryan N. Smith and Ioannis Rekleitis and Gregory Dudek},
title = {Adaptive exploration and sampling by heterogeneous robotic team},
booktitle = {Proc. of the IEEE International Conference on Robotics and Automation (ICRA 2018)},
year = {2018},
address = {Sydney, Australia}
} abstract: Physical sampling of water for off-site analysis is necessary for many applications like monitoring the quality of drinking water in reservoirs, understanding marine ecosystems, and measuring contamination levels in fresh-water systems. Robotic sampling enables to strategically collect water samples based on real-time measurements of physical and chemical properties gathered with onboard sensors. In this paper, we present a multi-robot, data-driven, watersampling strategy, where autonomous surface vehicles plan and execute@inproceedings{hansen2018autonomous,
title={Autonomous Marine Sampling Enhanced by Strategically Deployed Drifters in Marine Flow Fields},
author={Hansen, J. and Manjanna, S. and Li, A. Q. and Rekleitis, I. and Dudek, G.},
booktitle={OCEANS 2018 MTS/IEEE Charleston},
pages={1--7},
year={2018}
} abstract: We present a transportable system for ocean observations in which a small autonomous surface vehicle (ASV) adaptively collects spatially diverse samples with aid from a team of inexpensive, passive floating sensors known as drifters. Drifters can provide an increase in spatial coverage at little cost as they are propelled about the survey area by the ambient flow field instead of with actuators. Our iterative planning approach demonstrates how we can use the ASV to strategically deploy drifters into points of the flow field for high expected@inproceedings{hansen2018coverage,
title={Coverage optimization with non-actuated, floating mobile sensors using iterative trajectory planning in marine flow fields},
author={Hansen, J. and Dudek, G.},
booktitle={2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
pages={1906--1912},
year={2018},
organization={IEEE}
} abstract: This paper considers a spatial coverage problem in which a network of passive floating sensors is used to collect samples in a body of water. We employ an iterative measurement and modeling scheme to incrementally deploy sensors so as to achieve spatial coverage, despite only controlling the initial sample point. Once deployed, sensors are moved about a survey area by ambient surface currents. We demonstrate our results in simulation on 40 different ocean flow fields and compare against several baselines. This work provides a@inproceedings{manderson2018gaze,
title={Gaze Selection for Enhanced Visual Odometry During Navigation},
author={Manderson, Travis and Holliday, Andrew and Dudek, Gregory},
booktitle={Proceedings of the Conference on Computer and Robot Vision (CRV 2018)},
pages={110--117},
year={2018},
organization={IEEE},
address={Toronto, Canada}
} abstract: We present an approach to enhancing visual odometry and Simultaneous Localization and Mapping (SLAM) in the context of robot navigation by actively modulating the gaze direction to enhance the quality of the odometric estimates that are returned. We focus on two quality factors: i) stability of the visual features, and ii) consistency of the visual features with respect to robot motion and the associated correspondence between frames. We assume that local texture measures are associated with underlying scene content and thus with the quality of@inproceedings{manderson2018gpu,
author = {Travis Manderson and Gregory Dudek},
title = {GPU-Assisted Learning on an Autonomous Marine Robot for Vision-Based Navigation and Image Understanding},
booktitle = {Proceedings of Oceans Conference and Exposition 2018},
year = {2018},
address = {Charleston, United States}
} abstract: We present a GPU-based integrated robotic platform that enables collision avoidance, navigation, and image understanding on a single underwater vehicle. The platform enables observational tasks such as coral reef health assessment by enabling simultaneous operation of multiple image analysis taskswhile navigating in close proximity to obstacles. The integration of a GPU allows us to leverage deep neural networks for collision avoidance and automated object detection and classification while a general purpose CPU processes@inproceedings{Manjanna2018,
author = {Manjanna, S. and Li, A. Q. and Smith, R. N. and Rekleitis, I. and Dudek, G.},
title = {Heterogeneous Multi-Robot System for Exploration and Strategic Water Sampling},
booktitle = {2018 IEEE International Conference on Robotics and Automation (ICRA)},
year = {2018},
month = {May},
pages = {1--8}
} abstract: Physical sampling of water for off-site analysis is necessary for many applications like monitoring the quality of drinking water in reservoirs, understanding marine ecosystems, and measuring contamination levels in fresh-water systems. In this paper, the focus is on algorithms for efficient measurement and sampling using a multi-robot, data-driven, water-sampling behavior, where autonomous surface vehicles plan and execute water sampling using the chlorophyll density as a cue for plankton-rich water samples. We use two@inproceedings{Shkurti2018,
author = {Florian Shkurti and Nikhil Kakodkar and Gregory Dudek},
title = {Model-Based Probabilistic Pursuit via Inverse Reinforcement Learning},
booktitle = {Proceedings of the IEEE International Conference on Robotics and Automation (ICRA 2018)},
year = {2018},
address = {Sydney, Australia},
month = {May}
} abstract: We address the integrated prediction, planning, and control problem that enables a single follower robot (the photographer) to quickly re-establish visual contact with a moving target (the subject) that has escaped the follower's field of view. We deal with this scenario, which reactive controllers are typically ill-equipped to handle, by making plausible predictions about the long-and short-term behavior of the target, and planning pursuit paths that will maximize the chance of seeing the target again. At the core of our pursuit method is the use@inproceedings{manderson2018navigation,
title={Navigation in the Service of Enhanced Pose Estimation},
author={Manderson, Travis and Cheng, Ran and Meger, David and Dudek, Gregory},
booktitle={Proceedings of the 2018 International Symposium on Experimental Robotics (ISER 2018)},
year={2018},
address={Buenos Aires, Argentina}
} abstract: This paper addresses robust vision-based odometry for underwater robotics by autonomously adjusting the robot trajectory in real-time to optimize the quality of ongoing visual feedback. It is well-known that accurate Visual Odometry (VO) depends on both the presence of sufficient smooth surfaces with manageable reflectance functions and The robotic vehicle used in this work is a fully-autonomous marine system that uses vision for collision avoidance, navigation, and image understanding on a single vehicle (Fig. 2). This vehicle is a variant of@article{hansen2018planning,
title={Planning in Dynamic Environments with Conditional Autoregressive Models},
author={Hansen, Johanna and Kastner, Kyle and Courville, Aaron and Dudek, Gregory},
journal={arXiv preprint arXiv:1811.10097},
year={2018}
} abstract: We demonstrate the use of conditional autoregressive generative models (van den Oord et al., 2016a) over a discrete latent space (van den Oord et al., 2017b) for forward planning with MCTS. In order to test this method, we introduce a new environment featuring varying difficulty levels, along with moving goals and obstacles. The combination of high-quality frame generation and classical planning approaches nearly matches true environment performance for our task, demonstrating the usefulness of this method for model-based@inproceedings{Manjanna2018,
author = {Manjanna, S. and van Hoof, H. and Dudek, G.},
title = {Reinforcement Learning with Non-uniform State Representations for Adaptive Search},
booktitle = {2018 IEEE International Symposium on Safety, Security, and Rescue Robotics (SSRR)},
year = {2018},
month = {August},
pages = {1--7}
} abstract: R(τ) = exp(−T/c) with T is the time until the target is found if the target is found within H time steps, or 0 otherwise4. 3The reward map q is not normalized, yet after clearing a fraction g or probability mass, the probability that the target has not been found yet is 1 − g. If the target were not found yet, the normalized probability that the target Dudek, “Data-driven selective sampling for marine vehicles using multi-scale paths,” in IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Vancouver, Canada, September@inproceedings{holliday2018scale,
title={Scale-Robust Localization Using General Object Landmarks},
author={Holliday, A. and Dudek, G.},
booktitle={2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
pages={1688--1694},
year={2018},
organization={IEEE}
} abstract: Visual localization under large changes in scale is an important capability in many robotic mapping applications, such as localizing at low altitudes in maps built at high altitudes, or performing loop closure over long distances. Existing approaches, however, are robust only up to about a 3× difference in scale between map and query images. We propose a novel combination of deep-learning-based object features and state-of-the-art SIFT point-features that yields improved robustness to scale change. This technique is training-free and class@inproceedings{Li2018,
author = {Jimmy Li and Zhaoqi Xu and David Meger and Gregory Dudek},
title = {Semantic Scene Models for Visual Localization Under Large Viewpoint Changes},
booktitle = {Proceedings of the 15th Conference on Computer and Robot Vision (CRV 2018)},
year = {2018},
address = {Toronto},
month = {May}
} abstract: We propose an approach for camera pose estimation under large viewpoint changes using only 2D RGB images. This enables a mobile robot to relocalize itself with respect to a previously-visited scene when seeing it again from a completely new vantage point. In order to overcome large appearance changes, we integrate a variety of cues, including object detections, vanishing points, structure from motion, and object-to-object context in order to constrain the camera geometry, while simultaneously estimating the 3D pose of covisible@inproceedings{higuera2018synthesizing,
title={Synthesizing Neural Network Controllers with Probabilistic Model-Based Reinforcement Learning},
author={Higuera, Juan Camilo Gamboa and Meger, David and Dudek, Gregory},
booktitle={2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
pages={2538--2544},
year={2018},
organization={IEEE}
} abstract: We present an algorithm for rapidly learning neural network policies for robotics systems. The algorithm follows the model-based reinforcement learning paradigm and improves upon existing algorithms: PILeO and a sample-based version of PILeo with neural network dynamics (Deep-PILeO). To improve convergence, we propose a model-based algorithm that uses fixed random numbers and clips gradients during optimization. We propose training a neural network dynamics model using variational dropout with truncated Log@inproceedings{koreitem2018synthetically,
title={Synthetically trained 3d visual tracker of underwater vehicles},
author={Koreitem, K. and Li, J. and Karp, I. and Manderson, T. and Shkurti, F. and Dudek, G.},
booktitle={OCEANS 2018 MTS/IEEE Charleston},
pages={1--7},
year={2018},
organization={IEEE}
} abstract: We present a method for visually detecting and tracking the 3D pose of autonomous underwater vehicles, which aims to enable robust multi-robot convoying. We follow the approach of tracking-by-detection, which combines the robust, drift-free nature of object detection with the temporal consistency of tracking algorithms. Central to our method is a multi-output convolutional network that jointly predicts whether the target robot is present in the image (classification), the 2D bounding box around the target in the image plane, and@inproceedings{manderson2018vision,
title={Vision-based Autonomous Underwater Swimming in Dense Coral for Combined Collision Avoidance and Target Selection},
author={Manderson, Travis and Gamboa Higuera, Juan Camilo and Cheng, Ran and Dudek, Gregory},
booktitle={2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
pages={2018--2025},
year={2018},
organization={IEEE}
} abstract: We address the problem of learning vision-based, collision-avoiding, and target-selecting controllers in 3D, specifically in underwater environments densely populated with coral reefs. Using a highly maneuverable, dynamic, six-legged (or flippered) vehicle to swim underwater, we exploit real time visual feedback to make close-range navigation decisions that would be hard to achieve with other sensors. Our approach uses computer vision as the sole mechanism for both collision avoidance and visual target selection. In particular, we@article{sattar2018visual,
abstract = {We present an algorithm for underwater robots to visually detect and track human motion. Our objective is to enable human–robot interaction by allowing a robot to follow behind a human moving in (up to) six degrees of freedom. In particular, we have developed a system to allow a robot to detect, track and follow a scuba diver by using frequency-domain detection of biological motion patterns. The motion of biological entities is characterized by combinations of periodic motions which are inherently distinctive. This is especially true of},
author = {Sattar, Junaed and Dudek, Gregory},
journal = {Autonomous Robots},
pages = {111--124},
pub_year = {2018},
publisher = {Springer},
title = {Visual identification of biological motion for underwater human--robot interaction},
venue = {Autonomous Robots},
volume = {42}
} abstract: We present an algorithm for underwater robots to visually detect and track human motion. Our objective is to enable human–robot interaction by allowing a robot to follow behind a human moving in (up to) six degrees of freedom. In particular, we have developed a system to allow a robot to detect, track and follow a scuba diver by using frequency-domain detection of biological motion patterns. The motion of biological entities is characterized by combinations of periodic motions which are inherently distinctive. This is especially true of@inproceedings{Higuera2017,
author = {J. C. G. Higuera and D. Meger and G. Dudek},
title = {Adapting learned robotics behaviours through policy adjustment},
booktitle = {IEEE International Conference on Robotics and Automation (ICRA)},
year = {2017},
pages = {5837--5843}
} abstract: We present an approach to learning control policies for physical robots that achieves high efficiency by adjusting existing policies that have been learned on similar source systems, such as a similar robot with different physical parameters, or an approximate dynamics model simulator. This can be viewed as calibrating a policy learned on a source system, to match a desired behaviour in similar target systems. Our approach assumes that the trajectories described by the source robot are feasible on the target robot. By making this@article{Henderson2017,
author = {P. Henderson and W.-D. Chang and F. Shkurti and J. Hansen and D. Meger and G. Dudek},
title = {Benchmark environments for multitask learning in continuous domains},
year = {2017},
journal = {IROS},
} abstract: As demand drives systems to generalize to various domains and problems, the study of multitask, transfer and lifelong learning has become an increasingly important pursuit. In discrete domains, performance on the Atari game suite has emerged as the de facto benchmark for assessing multitask learning. However, in continuous domains there is a lack of agreement on standard multitask evaluation environments which makes it difficult to compare different approaches fairly. In this work, we describe a benchmark set of tasks that@inproceedings{li2017context,
title={Context-coherent scenes of objects for camera pose estimation},
author={Li, J. and Meger, D. and Dudek, G.},
booktitle={2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
pages={},
year={2017},
month={September},
organization={IEEE}
} abstract: We propose an approach to vision-based pose estimation using object recognition and identity. Whereas feature based scene recognition and pose estimation methods are well established as effective means for estimating motion and recognizing locations, feature-based methods depend critically on the detection of common local features from one view of a scene to another. We focus on place recognition and pose change estimation in the context of large changes in viewing position, even to the extent that no common surfaces are@article{st2017control,
abstract = {Due to the recent technological progress, Human–RobotInteraction (HRI) has become a major field of research in both engineering and artistic realms, particularly so in the last decade. The mainstream interests are, however, extremely diverse: challenges are continuously shifting, the evolution of robot'skills, as well as the advances in methods for understanding their environment radically impact the design and implementation of research prototypes. When directly deployed in a public installation or artistic performances, robots},
author = {St-Onge, David and Br{\`e}ches, Pierre-Yves and Sharf, Inna and Reeves, Nicolas and Rekleitis, Ioannis and Abouzakhm, Patrick and Girdhar, Yogesh and Harmat, Adam and Dudek, Gregory and Gigu{\`e}re, Philippe},
journal = {Robotics and Autonomous Systems},
pages = {165--186},
pub_year = {2017},
publisher = {Elsevier},
title = {Control, localization and human interaction with an autonomous lighter-than-air performer},
venue = {Robotics and …},
volume = {88}
} abstract: Due to the recent technological progress, Human–RobotInteraction (HRI) has become a major field of research in both engineering and artistic realms, particularly so in the last decade. The mainstream interests are, however, extremely diverse: challenges are continuously shifting, the evolution of robot'skills, as well as the advances in methods for understanding their environment radically impact the design and implementation of research prototypes. When directly deployed in a public installation or artistic performances, robots@inproceedings{manjanna2017data,
author = {S. Manjanna and G. Dudek},
title = {Data-driven selective sampling for marine vehicles using multi-scale paths},
booktitle = {2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
year = {2017},
month = {September}
} abstract: This paper addresses adaptive coverage of a spatial field without prior knowledge. Our application in this paper is to cover a region of the sea surface using a robotic boat, although the algorithmic approach has wider applicability. We propose an anytime planning technique for efficient data gathering using point-sampling based on non-uniform data-driven coverage. Our goal is to sense a particular region of interest in the environment and be able to reconstruct the measured spatial field. Since there are autonomous agents@inproceedings{Higuera2017,
author = {J. C. G. Higuera and D. Meger and G. Dudek},
title = {From simulation to the field: Learning to swim with the aqua robot},
booktitle = {ROSCon 2017},
year = {2017},
month = sep
} abstract: From simulation to the field: Learning to swim with the aqua robot@inproceedings{kalmbach2017learning,
title={Learning Seasonal Phytoplankton Communities with Topic Models},
author={Kalmbach, A. and Sosik, Heidi M. and Dudek, Gregory and Girdhar, Yogesh},
booktitle={IEEE Oceans},
year={2017},
note={Award winner},
eprint={1711.09013},
archivePrefix={arXiv},
primaryClass={cs.LG}
} abstract: In this work we develop and demonstrate a probabilistic generative model for phytoplankton communities. The proposed model takes counts of a set of phytoplankton taxa in a timeseries as its training data, and models communities by learning sparse co-occurrence structure between the taxa. Our model is probabilistic, where communities are represented by probability distributions over the species, and each time-step is represented by a probability distribution over the communities. The proposed approach uses a non@inproceedings{kalmbach2017phytoplankton,
title={Phytoplankton hotspot prediction with an unsupervised spatial community model},
author={Kalmbach, A. and Girdhar, Yogesh and Sosik, Heidi M. and Dudek, Gregory},
booktitle={IEEE International Conference on Robotics and Automation (ICRA)},
year={2017},
pages={2017},
note={arXiv preprint arXiv:1703.07309}
} abstract: Many interesting natural phenomena are sparsely distributed and discrete. Locating the hotspots of such sparsely distributed phenomena is often difficult because their density gradient is likely to be very noisy. We present a novel approach to this search problem, where we model the co-occurrence relations between a robot's observations with a Bayesian nonparametric topic model. This approach makes it possible to produce a robust estimate of the spatial distribution of the target, even in the absence of direct target@article{manderson2017robotic,
abstract = {This paper presents a system capable of autonomous surveillance and analysis of coral reef ecosystems using natural lighting. We describe our strategy to safely and effectively deploy a small marine robot to inspect a reef using its digital cameras. Image analysis using a (RBF‐SVM) radial basis function‐support vector machines in combination with (LBP) local binary pattern, Gabor and Hue descriptors developed in this work are able to analyze the resulting image data automatically and reliably by learning from the annotations of expert marine},
author = {Manderson, Travis and Li, Jimmy and Dudek, Natasha and Meger, David and Dudek, Gregory},
journal = {Journal of Field Robotics},
number = {1},
pages = {170--187},
pub_year = {2017},
publisher = {Wiley Online Library},
title = {Robotic coral reef health assessment using automated image analysis},
venue = {Journal of Field …},
volume = {34}
} abstract: This paper presents a system capable of autonomous surveillance and analysis of coral reef ecosystems using natural lighting. We describe our strategy to safely and effectively deploy a small marine robot to inspect a reef using its digital cameras. Image analysis using a (RBF‐SVM) radial basis function‐support vector machines in combination with (LBP) local binary pattern, Gabor and Hue descriptors developed in this work are able to analyze the resulting image data automatically and reliably by learning from the annotations of expert marine@inproceedings{shkurti2017topologically,
author = {F. Shkurti and G. Dudek},
title = {Topologically distinct trajectory predictions for probabilistic pursuit},
booktitle = {2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
month = {September},
year = {2017}
} abstract: We address the integrated planning and control problem that enables a single follower robot (the “photographer”) to maintain a moving target (the “subject”) in its field of view for as long as possible. We propose a real-time pursuit algorithm that seamlessly handles the often neglected, yet unavoidable, scenario in which the target escapes the follower's field of view; a scenario that simple, reactive controllers are ill-equipped to handle. Our algorithm aims to minimize the expected time until visual contact is re-established, which enables the@inproceedings{Shkurti2017,
author = {F. Shkurti and W.-D. Chang and P. Henderson and M. J. Islam and J. C. G. Higuera and J. Li and T. Manderson and A. Xu and G. Dudek and J. Sattar},
title = {Underwater multi-robot convoying using visual tracking by detection},
booktitle = {2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
year = {2017},
month = {September},
doi = {10.1109/IROS.2017.8202133},
note = {Also available as arXiv preprint arXiv:1709.08292}
} abstract: We present a robust multi-robot convoying approach that relies on visual detection of the leading agent, thus enabling target following in unstructured 3-D environments. Our method is based on the idea of tracking-by-detection, which interleaves efficient model-based object detection with temporal filtering of image-based bounding box estimation. This approach has the important advantage of mitigating tracking drift (ie drifting away from the target object), which is a common symptom of model-free trackers and is detrimental to sustained@inproceedings{quattrini2016data,
title={Data correlation and comparison from multiple sensors over a coral reef with a team of heterogeneous aquatic robots},
author={Quattrini Li, Adriano and Rekleitis, Ioannis and Manjanna, Santhosh and Kakodkar, Nikhil and Hansen, John and Dudek, Gregory and Bobadilla, Leonardo and Anderson, John and Smith, Ryan},
booktitle={International Symposium of Experimental Robotics (ISER)},
year={2016}
} abstract: This paper presents experimental insights from the deployment of an ensemble of heterogeneous autonomous sensor systems over a shallow coral reef. Visual, inertial, GPS, and ultrasonic data collected are compared and correlated to produce a comprehensive view of the health of the coral reef. Coverage strategies are discussed with a focus on the use of informed decisions to maximize the information collected during a fixed period of time.@inproceedings{Manjanna2016,
author = {S. Manjanna and N. Kakodkar and M. Meghjani and G. Dudek},
title = {Efficient terrain driven coral coverage using gaussian processes for mosaic synthesis},
booktitle = {2016 13th Conference on Computer and Robot Vision (CRV)},
year = {2016},
pages = {448--455},
publisher = {IEEE}
} abstract: In this paper we present an efficient method for visual mapping of open water environments using exploration and reward identification followed by selective visual coverage. In particular, we consider the problem of visual mapping a shallow water coral reef to provide an environmental assay. Our approach has two stages based on two classes of sensors: bathymetric mapping and visual mapping. We use a robotic boat to collect bathymetric data using a sonar sensor for the first stage and video data using a visual sensor for the second@inproceedings{Meghjani2016,
author = {Meghjani, Maithilee and Dudek, Gregory},
title = {Fast and efficient rendezvous in street networks},
booktitle = {2016 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
year = {2016},
pages = {1887--1893},
publisher = {IEEE}
} abstract: Fast and efficient rendezvous in street networks@inproceedings{li2016learning,
title={Learning to generalize 3d spatial relationships},
author={Li, Jimmy and Meger, D. and Dudek, G.},
booktitle={Robotics and Automation (ICRA), 2016 IEEE International Conference on},
pages={5744--5749},
year={2016},
organization={IEEE}
} abstract: This paper presents an approach to learn meaningful spatial relationships in an unsupervised fashion from the distribution of 3D object poses in the real world. Our approach begins by extracting an over-complete set of features to describe the relative geometry of two objects. Each relationship type is modeled using a relevance-weighted distance over this feature space. This effectively ignores irrelevant feature dimensions. Our algorithm RANSEM for determining subsets of data that share a relationship as well as the@inproceedings{xu2016maintaining,
title={Maintaining efficient collaboration with trust-seeking robots},
author={Xu, A. and Dudek, G.},
booktitle={Intelligent Robots and Systems (IROS), 2016 IEEE/RSJ International Conference on},
pages={3312--3319},
year={2016},
organization={IEEE}
} abstract: In this work, we grant robot agents the capacity to sense and react to their human supervisor's changing trust state, as a means to maintain the efficiency of their collaboration. We propose the novel formulation of Trust-Aware Conservative Control (TACtiC), in which the agent alters its behaviors momentarily whenever the human loses trust. This trust-seeking robot framework builds upon an online trust inference engine and also incorporates an interactive behavior adaptation technique. We present end-to-end instantiations of trust@article{girdhar2016modeling,
abstract = {This paper presents a novel approach to modeling curiosity in a mobile robot, which is useful for monitoring and adaptive data collection tasks, especially in the context of long term autonomous missions where pre-programmed missions are likely to have limited utility. We use a realtime topic modeling technique to build a semantic perception model of the environment, using which, we plan a path through the locations in the world with high semantic information content. The life-long learning behavior of the proposed perception},
author = {Girdhar, Yogesh and Dudek, Gregory},
journal = {Autonomous Robots},
pages = {1267--1278},
pub_year = {2016},
publisher = {Springer},
title = {Modeling curiosity in a mobile robot for long-term autonomous exploration and monitoring},
venue = {Autonomous Robots},
volume = {40}
} abstract: This paper presents a novel approach to modeling curiosity in a mobile robot, which is useful for monitoring and adaptive data collection tasks, especially in the context of long term autonomous missions where pre-programmed missions are likely to have limited utility. We use a realtime topic modeling technique to build a semantic perception model of the environment, using which, we plan a path through the locations in the world with high semantic information content. The life-long learning behavior of the proposed perception@inproceedings{Meghjani2016,
author = {Meghjani, Maithilee and Dudek, Gregory},
title = {Multi-target rendezvous search},
booktitle = {Proc. International Conference on Intelligent Robots and Systems (IROS)},
year = {2016},
pages = {2596--2603},
publisher = {IEEE},
note = {Nominee for best-paper award in the Search and Rescue category}
} abstract: In this paper, we examine multi-target search, where one or more targets must be found by a moving robot. Given the target's initial probability distribution or the expected search region, we present an analysis of three search strategies-Global maxima search, Local maxima search, and Spiral search. We aim at minimizing the mean-time-to-find and maximizing the total probability of finding the target. This leads to two types of illustrative performance metrics: minimum time capture and guaranteed capture. We validate the search strategies@inproceedings{Meghjani2016,
author = {Meghjani, M. and Manjanna, S. and Dudek, G.},
title = {Multi-target search strategies},
booktitle = {Proc. IEEE International Symposium on Safety, Security, and Rescue Robotics (SSRR)},
year = {2016},
pages = {328--333},
publisher = {IEEE},
note = {Finalist for best paper award}
} abstract: This paper addresses the problem of searching multiple non-adversarial targets using a mobile searcher in an obstacle-free environment. In practice, we are particularly interested in marine applications where the targets drift on the ocean surface. These targets can be surface sensors used for marine environmental monitoring, drifting debris, or lost divers in open water. Searching for a floating target requires prior knowledge about the search region and an estimate of the target's motion. This task becomes challenging when searching for@inproceedings{koreitem2016subsea,
title={Subsea fauna enumeration using vision-based marine robots},
author={Koreitem, K. and Girdhar, Y. and Cho, W. and Singh, H. and Pineda, J. and Dudek, G.},
booktitle={2016 13th Conference on Computer and Robot Vision (CRV)},
pages={101--108},
year={2016},
organization={IEEE}
} abstract: This paper describes a robotics system for population density estimation of marine organisms and vision-based algorithm for computing the associated population estimates. We focus on benthic fauna, through the use of Seabed AUV to collect benthic imagery, and then employ a support vector machine (SVM) for automated analysis of these images to estimate the population of the fauna of interest. The proposed approach is a significant improvement over existing techniques such as trawling, or manual inspection of images@inproceedings{manderson2016texture,
title={Texture-aware slam using stereo imagery and inertial information},
author={Manderson, T. and Shkurti, F. and Dudek, G.},
booktitle={2016 13th Conference on Computer and Robot Vision (CRV)},
pages={456--463},
year={2016},
organization={IEEE}
} abstract: We present a gaze control method that augments an existing stereo and inertial Simultaneous Localization And Mapping (SLAM) system by directing the stereo camera towards feature-rich regions of the scene. Our integrated active SLAM system is based on careful triangulation of visual features, existing successful nonlinear optimization, and visual loop closing frameworks. It relies on the tight coupling of IMU measurements with constraints imposed by visual correspondences from both stereo and motion. Alongside the SLAM@inproceedings{xu2016towards,
abstract = {We are interested in enhancing the efficiency of human–robot collaborations, especially in “supervisor-worker” settings where autonomous robots work under the supervision of a human operator. We believe that trust serves a critical role in modeling the interactions within these teams, and also in streamlining their efficiency. We propose an operational formulation of human–robot trust on a short interaction time scale, which is tailored to a practical tele-robotics setting. We also report on a controlled user study that collected},
author = {Xu, Anqi and Dudek, Gregory},
booktitle = {Robotics Research: The 16th International Symposium ISRR},
organization = {Springer},
pages = {113--129},
pub_year = {2016},
title = {Towards modeling real-time trust in asymmetric human--robot collaborations},
venue = {Robotics Research: The 16th International Symposium …}
} abstract: We are interested in enhancing the efficiency of human–robot collaborations, especially in “supervisor-worker” settings where autonomous robots work under the supervision of a human operator. We believe that trust serves a critical role in modeling the interactions within these teams, and also in streamlining their efficiency. We propose an operational formulation of human–robot trust on a short interaction time scale, which is tailored to a practical tele-robotics setting. We also report on a controlled user study that collected@inproceedings{St-Onge2015,
author = {David St-Onge and Nicolas Reeves and Philippe Gigu\`ere and Inna Sharf and Gregory Dudek and Ioannis Rekleitis and Pierre-Yves Br\`eches and Patrick Abouzakhm and Philippe Babin},
title = {AEROSTABILES: A new approach to HRI research},
booktitle = {Proceedings of the Tenth Annual ACM/IEEE International Conference on Human-Robot Interaction (Extended Abstracts)},
year = {2015},
pages = {277--277},
month = {March}
} abstract: Initiated as a research-creation project by professor and artist Nicolas Reeves, the Aerostabile project quickly expanded to include researchers and artists from a wide range of disciplines. Its current phase brings together four robotic and research-creation labs with various expertises in unstable and dynamic environments. The first group, under the direction of professor Inna Sharf, is based at the department of mechanical engineering at University McGill. It works on control and modeling of autonomous blimps for satellite@inproceedings{manjanna2015autonomous,
title={Autonomous gait selection for energy efficient walking},
author={Manjanna, Sandeep and Dudek, Gregory},
booktitle={Proceedings of the IEEE International Conference on Robotics and Automation (ICRA '15)},
pages={5155--5162},
year={2015},
location={Seattle, USA},
month={May}
} abstract: In this paper, we investigate the question of how a legged robot can walk efficiently by taking advantage of its ability to alter its gait as a function of statistical (large-scale) terrain properties. One of the contributions of this paper is the algorithm to achieve real-time terrain identification and autonomous gait adaptation on a legged robot. We approach this problem by first classifying the terrains based on their proprioceptive responses and identifying the terrain in real-time. Then we choose an optimal gait to best suit the identified terrain type. We@inproceedings{meger2015learning,
author = {Meger, David and Gamboa Higuera, Juan C. and Xu, A. and Giguere, P. and Dudek, Gregory},
title = {Learning legged swimming gaits from experience},
booktitle = {Proceedings of the IEEE International Conference on Robotics and Automation (ICRA '15)},
pages = {5155--5162},
year = {2015},
address = {Seattle, USA},
note = {best paper nominee}
} abstract: We present an end-to-end framework for realizing fully automated gait learning for a complex underwater legged robot. Using this framework, we demonstrate that a hexapod flipper-propelled robot can learn task-specific control policies purely from experience data. Our method couples a state-of-the-art policy search technique with a family of periodic low-level controls that are well suited for underwater propulsion. We demonstrate the practical efficacy of tabula rasa learning, that is, learning without the use of any prior knowledge, of@inproceedings{xu2015optimo,
author = {Anqi Xu and Gregory Dudek},
title = {OPTIMo: Online Probabilistic Trust Inference Model for Asymmetric Human-Robot Collaborations},
booktitle = {Proceedings of the 10th ACM/IEEE International Conference on Human-Robot Interactions (HRI '15)},
pages = {7},
location = {Portland, USA},
month = {March},
year = {2015}
} abstract: We present OPTIMo: an Online Probabilistic Trust Inference Model for quantifying the degree of trust that a human supervisor has in an autonomous robot" worker". Represented as a Dynamic Bayesian Network, OPTIMo infers beliefs over the human's moment-to-moment latent trust states, based on the history of observed interaction experiences. A separate model instance is trained on each user's experiences, leading to an interpretable and personalized characterization of that operator's behaviors and attitudes. Using datasets@inproceedings{Rezanejad2015,
author = {Morteza Rezanejad and Babak Samari and Ioannis Rekleitis and Kaleem Siddiqi and Gregory Dudek},
title = {Robust Environment Mapping Using Flux Skeletons},
booktitle = {2015 IEEE/RSJ International Conference on Intelligent Robots and Systems},
year = {2015},
pages = {Sept 28 - Oct 03},
address = {Hamburg, Germany}
} abstract: We consider how to directly extract a road map (also known as a topological representation) of an initially-unknown 2-dimensional environment via an on-line procedure which robustly computes a retraction of its boundaries. While such approaches are well known for their theoretical elegance, computing such representations in practice is complicated when the data is sparse and noisy. In this paper we present the online construction of a topological map and the implementation of a control law for guiding the robot to the nearest unexplored@inproceedings{manderson2015autonomous,
title={Towards Autonomous Robotic Coral Reef Health Assessment},
author={Manderson, T. and Meger, D. and Li, J. and Cortes Poza, D. and Dudek, N. and Dudek, G.},
booktitle={Proceedings of Field and Service Robotics (FSR)},
address={Toronto, Canada},
year={2015},
month={June 24--26}
} abstract: This paper addresses the automated analysis of coral in shallow reef environments up to 90 ft deep. During a series of robotic ocean deployments, we have collected a data set of coral and non-coral imagery from four distinct reef locations. The data has been annotated by an experienced biologist and presented as a representative challenge for visual understanding techniques. We describe baseline techniques using texture and color features combined with classifiers for two vision sub-tasks: live coral image classification and live coral@inproceedings{xu2015towards,
author = {Anqi Xu and Gregory Dudek},
title = {Towards Efficient Collaborations with Trust-Seeking Adaptive Robots},
booktitle = {Proceedings of the 10th Human-Robot Interaction Pioneers Workshop (HRI Pioneers '15)},
year = {2015},
pages = {2},
address = {Portland, USA},
month = {March}
} abstract: We are interested in asymmetric human-robot teams, where a human supervisor occasionally takes over control to aid an autonomous robot in a given task. Our research aims to optimize team efficiency by improving the robot's task performance, decreasing the human's workload, and building trust in the team. We envision synergistic collaborations where the robot adapts its behaviors dynamically to optimize efficacy, reduce manual interventions, and actively seek for greater trust. We describe recent works that study two@inproceedings{zhang2015uncertainty,
author = {Q. Zhang and I. Rekleitis and G. Dudek},
title = {Uncertainty Reduction via Heuristic Search Planning on Hybrid Metric/Topological Map},
booktitle = {Proceedings of Conference on Computer and Robot Vision},
year = {2015},
address = {Halifax, NS},
month = {June}
} abstract: Uncertainty Reduction via Heuristic Search Planning on Hybrid Metric/Topological Map@inproceedings{Meger2014,
author = {David Meger and Florian Shkurti and David Cortes Poza and Philippe Giguere and Gregory Dudek},
title = {3D trajectory synthesis and control for a legged swimming robot},
booktitle = {Proc. IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2014)},
year = {2014},
address = {Chicago, IL},
month = {Sept.},
pages = {2257--2264}
} abstract: Inspection and exploration of complex underwater structures requires the development of agile and easy to program platforms. In this paper, we describe a system that enables the deployment of an autonomous underwater vehicle in 3D environments proximal to the ocean bottom. Unlike many previous approaches, our solution: uses oscillating hydrofoil propulsion; allows for stable control of the robot's motion and sensor directions; allows human operators to specify detailed trajectories in a natural fashion; and has been@inproceedings{xu2014adaptive,
title={Adaptive Parameter EXploration (APEX): Adaptation of robot autonomy from human participation},
author={Xu, Anqi and Kalmbach, Arnold and Dudek, Gregory},
booktitle={Proc. IEEE International Conference on Robotics and Automation (ICRA 2014)},
pages={3315--3322},
year={2014},
location={Hong Kong}
} abstract: The problem of Adaptation from Participation (AfP) aims to improve the efficiency of a human-robot team by adapting a robot's autonomous systems and behaviors based on command-level input from a human supervisor. As a solution to AfP, the Adaptive Parameter EXploration (APEX) algorithm continuously explores the space of all possible parameter configurations for the robot's autonomous system in an online and anytime manner. Guided by information deduced from the human's latest intervening commands, APEX is capable of@inproceedings{Meghjani2014,
author = {Malika Meghjani and Florian Shkurti and Juan Camilo Gamboa Higuera and Arnold Kalmbach and David Whitney and Gregory Dudek},
title = {Asymmetric Rendezvous Search at Sea},
booktitle = {Proc. Conference on Computer and Robot Vision (CRV 2014)},
pages = {7},
address = {Montreal, Canada},
month = {June},
year = {2014}
} abstract: In this paper we address the rendezvous problem between an autonomous underwater vehicle (AUV) and a passively floating drifter on the sea surface. The AUV's mission is to keep an estimate of the floating drifter's position while exploring the underwater environment and periodically attempting to rendezvous with it. We are interested in the case where the AUV loses track of the drifter, predicts its location and searches for it in the vicinity of the predicted location. We parameterize this search problem with respect to both the uncertainty@article{girdhar2014autonomous,
abstract = {The exploration of dangerous environments such as underwater coral reefs and shipwrecks is a difficult and potentially life-threatening task for humans, which naturally makes the use of an autonomous robotic system very appealing. This paper presents such an autonomous system, which is capable of autonomous exploration, and shows its use in a series of experiments to collect image data in challenging underwater marine environments. We present novel contributions on three fronts. First, we present an online topic-modeling-based},
author = {Girdhar, Yogesh and Giguere, Philippe and Dudek, Gregory},
journal = {The International Journal of Robotics Research},
number = {4},
pages = {645--657},
pub_year = {2014},
publisher = {SAGE Publications Sage UK: London, England},
title = {Autonomous adaptive exploration using realtime online spatiotemporal topic modeling},
venue = {The International Journal of …},
volume = {33}
} abstract: The exploration of dangerous environments such as underwater coral reefs and shipwrecks is a difficult and potentially life-threatening task for humans, which naturally makes the use of an autonomous robotic system very appealing. This paper presents such an autonomous system, which is capable of autonomous exploration, and shows its use in a series of experiments to collect image data in challenging underwater marine environments. We present novel contributions on three fronts. First, we present an online topic-modeling-based@inproceedings{Girdhar2014,
author = {Yogesh Girdhar and David Whitney and Gregory Dudek},
title = {Curiosity Based Exploration for Learning Terrain Models},
booktitle = {Proc. IEEE International Conference on Robotics and Automation (ICRA)},
year = {2014},
pages = {7},
address = {Hong Kong},
month = {May},
eprint = {arXiv:1310.6767}
} abstract: We present a robotic exploration technique in which the goal is to learn a visual model that can be used to distinguish between different terrains and other visual components in an unknown environment. We use ROST, a realtime online spatiotemporal topic modeling framework to model these terrains using the observations made by the robot, and then use an information theoretic path planning technique to define the exploration path. We conduct experiments with aerial view and underwater datasets with millions of observations and@inproceedings{Girdhar2014,
author = {Yogesh Girdhar and Gregory Dudek},
title = {Exploring Underwater Environments with Curiosity},
booktitle = {Proc. Conference on Computer and Robot Vision (CRV 2014)},
pages = {7 pages},
address = {Montreal, Canada},
month = {June},
year = {2014}
} abstract: This paper presents a novel approach to modeling curiosity in a mobile robot, which is useful for monitoring and adaptive data collection tasks. We use ROST, a real time topic modeling framework to build a semantic perception model of the environment, using which, we plan a path through the locations in the world with high semantic information content. We demonstrate the approach using the Aqua robot in a variety of different scenarios, and find the robot be able to do tasks such as coral reef inspection, diver following, and sea floor@inproceedings{shkurti2014maximizing,
author = {Florian Shkurti and Gregory Dudek},
title = {Maximizing visibility in collaborative trajectory planning},
booktitle = {Proc. IEEE International Conference on Robotics and Automation (ICRA)},
year = {2014},
pages = {3315--3322},
address = {Hong Kong},
month = {May}
} abstract: In this paper we address the issue of coordinating the trajectories of two collaborating robots in environments with obstacles so that visibility between them is maximized in the presence of competing constraints. Specifically, we examine the problem of allowing one robot (the “photographer”) to follow another robot (“the subject”) through a planar environment while maintaining visual contact to the maximum degree consistent with an efficient traversal. This problem has numerous applications, for instance in scenarios where communication@inproceedings{meghjani2014multi,
author = {Malika Meghjani and Gregory Dudek},
title = {Multi-agent rendezvous on street networks},
booktitle = {Proc. IEEE International Conference on Robotics and Automation (ICRA 2014)},
pages = {5792--5797},
year = {2014},
address = {Hong Kong},
month = {May}
} abstract: In this paper we present an algorithm for finding a distance optimal rendezvous location with respect to both initial and target locations of the mobile agents. These agents can be humans or robots, who need to meet and split while performing a collaborative task. Our aim is to embed the meeting process within a background activity such that the agents travel through the rendezvous location while taking the shortest paths to their respective target locations. We analyze this problem in a street network scenario with two agents who are@article{dudek2014special,
abstract = {This issue of Autonomous Robots presents journal articles that are based on papers originally presented at the 2013 Robotics Science and Systems conference, held in Berlin, Germany. Although these were selected by a committee to exemplify the best papers presented that year, the decision over which papers to include was a difficult one due to the substantial number of very strong papers in the cohort. While the strength of these papers can be partially attributed to the fact that the conference itself generally attracts strong},
author = {Dudek, Gregory and Fox, Dieter},
journal = {Autonomous Robots},
pages = {333--334},
pub_year = {2014},
publisher = {Springer},
title = {Special issue on robotics: science and systems},
venue = {Autonomous Robots},
volume = {37}
} abstract: This issue of Autonomous Robots presents journal articles that are based on papers originally presented at the 2013 Robotics Science and Systems conference, held in Berlin, Germany. Although these were selected by a committee to exemplify the best papers presented that year, the decision over which papers to include was a difficult one due to the substantial number of very strong papers in the cohort. While the strength of these papers can be partially attributed to the fact that the conference itself generally attracts strong@inproceedings{gamboa2013fair,
author = {Juan Camilo Gamboa Higuera and Gregory Dudek},
title = {Fair Subdivision of Multirobot Tasks},
booktitle = {Proceedings of the IEEE International Conference on Robotics and Automation (ICRA)},
year = {2013},
pages = {6},
address = {Karlsruhe, Germany},
month = {May}
} abstract: We study the problem of distributing a single global task between a group of heterogeneous robots. We view this problem as a fair division game. In this setting, every robot defines a preference function over parts of the task according to its sensing and motion capabilities. These preferences are described by density functions over the task. With such interpretation, we want to find an allocation of the global task that maximizes the probability of task completion. We first formulate the task distribution problem as a fair subdivision problem and@inproceedings{dey2013ninja,
title={Ninja Legs: Amphibious One Degree of Freedom Robotic Legs},
author={Dey, Bir Bikram and Manjanna, Sandeep and Dudek, Gregory},
booktitle={Proceedings of the 2013 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS '13)},
year={2013},
address={Tokyo, Japan},
month={November}
} abstract: In this paper we propose a design of a class of robotic legs (known as “Ninja legs”) that enable amphibious operation, both walking and swimming, for use on a class of hexapod robots. Amphibious legs equip the robot with a capability to explore diverse locations in the world encompassing both those that are on the ground as well as underwater. In this paper we work with a hexapod robot of the Aqua vehicle family (based on a body plan first developed by Buehler et al.[1]), which is an amphibious robot that employs legs for@inproceedings{Shkurti2013,
author = {Florian Shkurti and Gregory Dudek},
title = {On the complexity of searching for an evader with a faster pursuer},
booktitle = {Proceedings of the IEEE International Conference on Robotics and Automation (ICRA)},
year = {2013},
pages = {6},
address = {Karlsruhe, Germany},
month = {May}
} abstract: In this paper we examine pursuit-evasion games in which the pursuer has higher speed than the evader. This scenario is motivated by visibility-based pursuit-evasion problems, particularly by the question of what happens when the pursuer loses visual track of the moving evader. In these cases the pursuer has two options for recovering visual contact with the evader: to perform search over the possible locations where the evader might be moving, or to clear the environment, in other words to progressively search it without@inproceedings{kalmbach2013unsupervised,
title={Unsupervised Environment Recognition and Modeling using Sound Sensing},
author={Kalmbach, Arnold and Girdhar, Yogesh and Dudek, Gregory},
booktitle={Proceedings of the IEEE International Conference on Robotics and Automation (ICRA)},
pages={6},
year={2013},
address={Karlsruhe, Germany},
month={May}
} abstract: We discuss the problem of automatically discovering different acoustic regions in the world, and then labeling the trajectory of a robot using these region labels. We use quantized Mel Frequency Cepstral Coefficients (MFCC) as low level features, and a temporally smoothed variant of Latent Dirichlet Allocation (LDA) to compute both the region models, and most likely region labels associated with each time step in the robot's trajectory. We validate our technique by showing results from two datasets containing sound recorded from 51 and 43@inproceedings{manjanna2013using,
title={Using Gait Change for Terrain Sensing by Robots},
author={Manjanna, Sandeep and Gigu{\`e}re, Philippe and Dudek, Gregory},
booktitle={Proceedings of the 10th International Conference on Computer and Robot Vision (CRV '13)},
pages={7},
year={2013},
address={Regina, Canada},
month={May}
} abstract: In this paper we examine the interplay between terrain classification accuracy and gait in a walking robot, and show how changes in walking speed can be used for terrain-dependent walk optimizations, as well as to enhance terrain identification. The details of a walking gait have a great influence on the performance of locomotive systems and their interaction with the terrain. Most legged robots can benefit from adapting their gait (and specifically walk speed) to the particular terrain on which they are walking. To achieve this, the agent should@inproceedings{Giguere2013,
author = {Philippe Gigu\`ere and Yogesh Girdhar and Gregory Dudek},
title = {Wide-Speed Autopilot System for a Swimming Hexapod Robot},
booktitle = {Proceedings of the 10th International Conference on Computer and Robot Vision (CRV '13)},
year = {2013},
pages = {7 pages},
address = {Regina, Canada},
month = {May},
url = {http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=6569178&isnumber=6569168}
} abstract: For underwater swimming robots, which use the unconventional method of oscillating flippers for propulsion and control, being able to move stably at various velocities is challenging. This stable motion facilitates navigation, avoids blurring the images taken by a camera motion, and enables longterm observations of specific locations. Previous experiments with our swimming robot Aqua have shown that its autopilot system must adapt the control parameters as a function of speed. The reason is that the dynamics of both the@inproceedings{Girdhar2013,
author = {Yogesh Girdhar and Philippe Giguere and Gregory Dudek},
title = {Autonomous Adaptive Underwater Exploration using Online Topic Modeling},
booktitle = {Experimental Robotics},
pages = {789--802},
year = {2013},
note = {Presented at the International Symposium on Experimental Robotics (ISER)}
} abstract: Exploration of underwater environments, such as coral reefs and ship wrecks, is a difficult and potentially dangerous tasks for humans, which naturally makes the use of an autonomous robotic system very appealing. This paper presents such an autonomous system, and shows its use in a series of experiments to collect image data in an underwater marine environment. We presents novel contributions on three fronts. First, we present an online topic-modeling based technique to describe what is being observed using a low@book{Desai2013,
editor = {Jaydev Desai and Gregory Dudek and Oussama Khatib and Vijay Kumar},
title = {Experimental Robotics: The 13th International Symposium on Experimental Robotics},
year = {2013},
publisher = {Springer},
doi = {10.1007/978-3-319-00065-7},
isbn = {978-3-319-00064-0},
isbn = {978-3-319-00065-7 (eBook)},
issn = {1610-742X (electronic)}
} abstract: The International Symposium on Experimental Robotics (ISER) is a series of bi-annual meetings, which are organized, in a rotating fashion around North America, Europe and Asia/Oceania. The goal of ISER is to provide a forum for research in robotics that focuses on novelty of theoretical contributions validated by experimental results. The meetings are conceived to bring together, in a small group setting, researchers from around the world who are in the forefront of experimental robotics research. This unique reference presents the@inproceedings{Girdhar2013,
author = {Yogesh Girdhar and Philippe Giguere and Gregory Dudek},
title = {Autonomous Adaptive Underwater Exploration using Online Topic Modeling},
booktitle = {Experimental Robotics},
pages = {789--802},
year = {2013},
note = {Presented at the International Symposium on Experimental Robotics (ISER)}
} abstract: Exploration of underwater environments, such as coral reefs and ship wrecks, is a difficult and potentially dangerous tasks for humans, which naturally makes the use of an autonomous robotic system very appealing. This paper presents such an autonomous system, and shows its use in a series of experiments to collect image data in an underwater marine environment. We presents novel contributions on three fronts. First, we present an online topic-modeling based technique to describe what is being observed using a low@inproceedings{Girdhar2012,
author = {Yogesh Girdhar and Gregory Dudek},
title = {Efficient on-line data summarization using extremum summaries},
booktitle = {Proceedings of the IEEE International Conference on Robotics and Automation (ICRA)},
year = {2012},
month = {May}
} abstract: We are interested in the task of online summarization of the data observed by a mobile robot, with the goal that these summaries could be then be used for applications such as surveillance, identifying samples to be collected by a planetary rover, and site inspections to detect anomalies. In this paper, we pose the summarization problem as an instance of the well known k-center problem, where the goal is to identify k observations so that the maximum distance of any observation from a summary sample is minimized. We focus on@inproceedings{Girdhar2012,
author = {Yogesh Girdhar and Anqi Xu and Florian Shkurti and Juan Camilo Gamboa Higuera and Malika Meghjani and Philippe Gigu\`ere and Ioannis Rekleitis and Gregory Dudek},
title = {Monitoring Marine Environments using a Team of Heterogeneous Robots},
booktitle = {RSS 2012 Workshop on Robotics for Environmental Monitoring (WREM 2012)},
year = {2012},
address = {Sydney, Australia}
} abstract: We present a novel approach for monitoring marine environments by a team of heterogeneous robots, comprising of a fixed-wing aerial vehicle, an autonomous airboat, and a legged underwater robot. The goal is to receive a region of interest from a remote human supervisor, and then using the coordinated effort of the robot team, produce a concise summary consisting of a small number of images, which capture the visual diversity of the region of interest. The summary could then be used by a human supervisor to plan for@inproceedings{Shkurti2012,
author = {Florian Shkurti and Anqi Xu and Malika Meghjani and Juan Camilo Gamboa Higuera and Yogesh Girdhar and Philippe Gigu\`ere and Bir Bikram Dey and Jimmy Li and Arnold Kalmbach and Chris Prahacs and Katrine Turgeon and Ioannis Rekleitis and Gregory Dudek},
title = {Multi-Domain Monitoring of Marine Environments Using a Heterogeneous Robot Team},
booktitle = {Proceedings of the 2012 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS '12)},
year = {2012},
month = {October},
address = {Algarve, Portugal}
} abstract: In this paper we describe a heterogeneous multi-robot system for assisting scientists in environmental monitoring tasks, such as the inspection of marine ecosystems. This team of robots is comprised of a fixed-wing aerial vehicle, an autonomous airboat, and an agile legged underwater robot. These robots interact with off-site scientists and operate in a hierarchical structure to autonomously collect visual footage of interesting underwater regions, from multiple scales and mediums. We discuss organizational and scheduling@inproceedings{Meghjani2012,
author = {Malika Meghjani and Gregory Dudek},
title = {Multi-Robot Exploration and Rendezvous on Graphs},
booktitle = {Proceedings of the 2012 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS '12)},
year = {2012},
address = {Algarve, Portugal},
month = {October}
} abstract: We address the problem of arranging a meeting (or rendezvous) between two or more robots in an unknown bounded topological environment, starting at unknown locations, without any communication. The goal is to rendezvous in minimum time such that the robots can share resources for performing any global task. We specifically consider a global exploration task executed by two or more robots. Each robot explores the environment simultaneously, for a specified time, then selects potential rendezvous locations, where it@inproceedings{GamboaHiguera2012,
author = {Juan Camilo Gamboa Higuera and Anqi Xu and Florian Shkurti and Gregory Dudek},
title = {Socially-Driven Collective Path Planning for Robot Missions},
booktitle = {Proceedings of the 9th Canadian Conference on Computer and Robot Vision (CRV '12)},
year = {2012},
pages = {417--424},
address = {Toronto, Canada},
month = {May}
} abstract: We address the problem of path planning for robot missions based on waypoints suggested by multiple human users. These users may be operating under distinct mission objectives and hence suggest different locations for the robot to visit. We formulate this problem using a constrained optimization approach by imposing various operational considerations, such as the robot's maximum traversable distance. We then propose an approximative path planning algorithm with parameterized control over the degree of" social fairness" in the selection of@misc{desai2014special,
abstract = {This special issue consists of 12 papers drawn from contributions at the Thirteenth International Symposium on Experimental Robotics in 2012 (ISER'12). ISER is a series of biennial symposia whose goal is to provide the robotics community with a forum for research driven by creative ideas, bold visions, new systems, and novel applications of robotics, emphasizing experimental work. The ISER tradition fosters scholarly work that either addresses validation of theoretical paradigms through careful experimentation or the},
author = {Desai, Jaydev P and Dudek, Gregory and Khatib, Oussama and Kumar, Vijay},
journal = {The International Journal of Robotics Research},
number = {4},
pages = {487--488},
pub_year = {2014},
publisher = {SAGE Publications Sage UK: London, England},
title = {Special Issue of the Thirteenth International Symposium on Experimental Robotics, 2012},
venue = {… International Journal of …},
volume = {33}
} abstract: This special issue consists of 12 papers drawn from contributions at the Thirteenth International Symposium on Experimental Robotics in 2012 (ISER'12). ISER is a series of biennial symposia whose goal is to provide the robotics community with a forum for research driven by creative ideas, bold visions, new systems, and novel applications of robotics, emphasizing experimental work. The ISER tradition fosters scholarly work that either addresses validation of theoretical paradigms through careful experimentation or the@inproceedings{xu2012trust,
title={Trust-Driven Interactive Visual Navigation for Autonomous Robots},
author={Xu, Anqi and Dudek, Gregory},
booktitle={Proceedings of the IEEE International Conference on Robotics and Automation (ICRA '12)},
pages={3922--3929},
year={2012},
organization={IEEE},
address={St. Paul, USA},
month={May}
} abstract: We describe a model of “trust” in human-robot systems that is inferred from their interactions, and inspired by similar concepts relating to trust among humans. This computable quantity allows a robot to estimate the extent to which its performance is consistent with a human's expectations, with respect to task demands. Our trust model drives an adaptive mechanism that dynamically adjusts the robot's autonomous behaviors, in order to improve the efficiency of the collaborative team. We illustrate this trust-driven methodology through an interactive@article{giguere2011simple,
abstract = {This paper describes a tactile probe designed for surface identification in a context of all-terrain low-velocity mobile robotics. The proposed tactile probe is made of a small metallic rod with a single-axis accelerometer attached near its tip. Surface identification is based on analyzing acceleration patterns induced at the tip of this mechanically robust tactile probe, while it is passively dragged along a surface. A training dataset was collected over ten different indoor and outdoor surfaces. Classification results for an artificial neural network},
author = {Giguere, Philippe and Dudek, Gregory},
journal = {IEEE Transactions on Robotics},
number = {3},
pages = {534--544},
pub_year = {2011},
publisher = {IEEE},
title = {A simple tactile probe for surface identification by mobile robots},
venue = {IEEE Transactions on Robotics},
volume = {27}
} abstract: This paper describes a tactile probe designed for surface identification in a context of all-terrain low-velocity mobile robotics. The proposed tactile probe is made of a small metallic rod with a single-axis accelerometer attached near its tip. Surface identification is based on analyzing acceleration patterns induced at the tip of this mechanically robust tactile probe, while it is passively dragged along a surface. A training dataset was collected over ten different indoor and outdoor surfaces. Classification results for an artificial neural network@incollection{Girdhar2011,
author = {Yogesh Girdhar and Gregory Dudek},
title = {A surprising problem in navigation},
booktitle = {Vision in 3D Environments},
publisher = {Cambridge University Press},
year = {2011},
pages = {228--252},
isbn = {9781107001756}
} abstract: A surprising problem in navigation@inproceedings{Meghjani2011,
author = {Malika Meghjani and Gregory Dudek},
title = {Combining Multi-Robot Exploration and Rendezvous},
booktitle = {Proceedings of the 8th Canadian Conference on Computer and Robot Vision (CRV '11)},
year = {2011},
pages = {80--85},
address = {St. John's, Newfoundland, Canada},
month = {May}
} abstract: We consider the problem of exploring an unknown environment with a pair of mobile robots. The goal is to make the robots meet (or rendezvous) in minimum time such that there is a maximum speed gain of the exploration task. The key challenge in achieving this goal is to rendezvous with the least possible dependency on communication. This single constraint involves several sub-problems: finding unique potential rendezvous locations in the environment, ranking these locations based on their uniqueness and synchronizing with the@inproceedings{virie2011conformative,
author = {Patrick Virie and Gregory Dudek},
title = {Conformative Filter: A Probabilistic Framework for Localization in Reduced Space},
booktitle = {Proceedings of the 8th Canadian Conference on Computer and Robot Vision (CRV '11)},
year = {2011},
pages = {24--31},
address = {St. John's, Newfoundland, Canada},
month = {May}
} abstract: Algorithmic problem reduction is a fundamental approach to problem solving in many fields, including robotics. To solve a problem using this scheme, we must reduce the problem into another one for which solutions exist. The reduction function, which infers a conformation between the problem and the solution space, plays an important role in solution evaluation and is sometimes used to transform the solutions into the problem domain. We consider robot path planning in the context of algorithmic problem reduction where a reduction can be@inproceedings{shkurti2011feature,
author = {Florian Shkurti and Ioannis Rekleitis and Gregory Dudek},
title = {Feature Tracking Evaluation for Pose Estimation in Underwater Environments},
booktitle = {Proceedings of the 8th Canadian Conference on Computer and Robot Vision (CRV '11)},
pages = {160--167},
year = {2011},
address = {St. John's, Newfoundland, Canada},
month = {May}
} abstract: In this paper we present the computer vision component of a 6DOF pose estimation algorithm to be used by an underwater robot. Our goal is to evaluate which feature trackers enable us to accurately estimate the 3D positions of features, as quickly as possible. To this end, we perform an evaluation of available detectors, descriptors, and matching schemes, over different underwater datasets. We are interested in identifying combinations in this search space that are suitable for use in structure from motion algorithms, and more@inproceedings{xu2011fourier,
author = {Anqi Xu and Gregory Dudek},
title = {Fourier Tag: A Smoothly Degradable Fiducial Marker System with Configurable Payload Capacity},
booktitle = {Proc. Conference on Computer and Robot Vision (CRV '11)},
year = {2011},
pages = {40--47},
address = {St. John's, Newfoundland, Canada},
month = {May}
} abstract: We describe the design and implementation of a fiducial marker system that encodes data in the frequency spectrum of a synthetic image. This distinctive approach to marker synthesis and data encoding allows for partial data extraction in adverse imaging conditions, and can significantly extend the detection range through graceful data degradation. Additional digital encoding and image construction techniques are used to increase the payload capacity, and also to store 3-D pose information in each fiducial marker. This fiducial marker scheme can@inproceedings{li2011graphical,
title={Graphical State Space Programming: A Visual Programming Paradigm for Robot Task Specification},
author={Li, Jimmy and Xu, Anqi and Dudek, Gregory},
booktitle={2011 IEEE International Conference on Robotics and Automation (ICRA)},
pages={4846--4853},
year={2011},
organization={IEEE},
address={Shanghai, China},
month={May}
} abstract: We describe a framework that combines a software development paradigm, a software visualization technique, and a tool for robot programming. This infrastructure is called" Graphical State Space Programming"(GSSP), and allows robot application programs to be decomposed and visualized within state-dependent views. Our approach simplifies and expedites the programming process for robot routines and behaviors, and we examine the performance improvement that ensues through a set of controlled user studies. The usability@inproceedings{Girdhar2011MARE,
author = {Yogesh Girdhar and Anqi Xu and Bir Bikram Dey and Malika Meghjani and Florian Shkurti and Ioannis Rekleitis and Gregory Dudek},
title = {MARE: Marine Autonomous Robotic Explorer},
booktitle = {Proceedings of the 2011 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS '11)},
year = {2011},
pages = {5048--5053},
address = {San Francisco, USA},
doi = {10.1109/IROS.2011.6048582}
} abstract: We present MARE, an autonomous airboat robot that is suitable for exploration-oriented tasks, such as inspection of coral reefs and shallow seabeds. The combination of this platform's particular mechanical properties and its powerful software framework enables it to function in a multitude of potential capacities, including autonomous surveillance, mapping, and search operations. In this paper we describe two different exploration strategies and their implementation using the MARE platform. First, we discuss the application of an@inproceedings{Girdhar2011,
author = {Yogesh Girdhar and Gregory Dudek},
title = {Offline Navigation Summaries},
booktitle = {Proceedings of the IEEE International Conference on Robotics and Automation (ICRA2011)},
year = {2011},
pages = {5769--5775},
address = {Shanghai, China},
month = {May 9--13}
} abstract: Our objective is to find a small set of images that summarize a robot's visual experience along a path. We present a novel on-line algorithm for this task. This algorithm is based on a new extension to the classical Secretaries Problem. We also present an extension to the idea of Bayesian Surprise, which we then use to measure the fitness of an image as a summary image.@inproceedings{Girdhar2011,
author = {Yogesh Girdhar and Gregory Dudek},
title = {Online Visual Vocabularies},
booktitle = {Proceedings of the 8th Canadian Conference on Computer and Robot Vision (CRV '11)},
year = {2011},
pages = {191--196},
address = {St. John's, Newfoundland, Canada},
month = {May}
} abstract: The idea of an online visual vocabulary is proposed. In contrast to the accepted strategy of generating vocabularies offline, using the k-means clustering over all the features extracted form all the images in a dataset, an online vocabulary is dynamic and evolves iteratively over time as new observations are made. Hence, it is much more suitable for online robotic applications, such as exploration, landmark detection, and SLAM, where the future is unknown. We present two different strategies for building online vocabularies. The first@inproceedings{xu2011optimal,
title={Optimal Complete Terrain Coverage using an Unmanned Aerial Vehicle},
author={Xu, Anqi and Viriyasuthee, Chatavut and Rekleitis, Ioannis and Dudek, Gregory},
booktitle={Proceedings of the 2011 IEEE International Conference on Robotics and Automation (ICRA'11)},
pages={2513--2519},
year={2011},
organization={IEEE},
address={Shanghai, China},
month={May}
} abstract: We present the adaptation of an optimal terrain coverage algorithm for the aerial robotics domain. The general strategy involves computing a trajectory through a known environment with obstacles that ensures complete coverage of the terrain while minimizing path repetition. We introduce a system that applies and extends this generic algorithm to achieve automated terrain coverage using an aerial vehicle. Ex tensive experimental results in simulation validate the presented system, along with data from over 100 kilometers of@article{Chen2011,
author = {G. Y. Chen and Gregory Dudek and L. A. Torres-Mendez},
title = {Scene reconstruction with sparse range data and intensity information},
journal = {Optical Engineering},
volume = {50},
number = {9},
pages = {097002},
year = {2011},
doi = {10.1117/1.3617455},
url = {http://dx.doi.org/10.1117/1.3617455}
} abstract: Scene reconstruction with sparse range data and intensity information@inproceedings{sattar2011towards,
title={Towards Quantitative Modeling of Task Confirmations in Human-Robot Dialog},
author={Sattar, Junaed and Dudek, Gregory},
booktitle={Proceedings of the IEEE International Conference on Robotics and Automation (ICRA)},
pages={1957--1963},
year={2011},
address={Shanghai, China},
month={May 9--13}
} abstract: We present a technique for robust human-robot interaction taking into consideration uncertainty in input and task execution costs incurred by the robot. Specifically, this research aims to quantitatively model confirmation feedback, as required by a robot while communicating with a human operator to perform a particular task. Our goal is to model human-robot interaction from the perspective of risk minimization, taking into account errors in communication," risk" involved in performing the required task, and task execution costs@inproceedings{xu2010vision,
title={A Vision-Based Boundary Following Framework for Aerial Vehicles},
author={Xu, Anqi and Dudek, Gregory},
booktitle={Proceedings of the 2010 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS '10)},
pages={81--86},
year={2010},
organization={IEEE},
address={Taipei, Taiwan},
month={October}
} abstract: We present an integration of classical computer vision techniques to achieve real-time autonomous steering of an unmanned aircraft along the boundary of different regions. Using an unified conceptual framework, we illustrate solutions for tracking coastlines and for following roads surrounded by forests. In particular, we exploit color and texture properties to differentiate between region types in the aforementioned domains. The performance of our system is evaluated using different experimental approaches, which includes a fully@inproceedings{sattar2010graphical,
author = {Junaed Sattar and Anqi Xu and Gabrielle Charette and Gregory Dudek},
title = {Graphical State-Space Programmability as a Natural Interface for Robotic Control},
booktitle = {Proceedings of the 2010 IEEE International Conference on Robotics and Automation (ICRA '10)},
year = {2010},
pages = {4609--4614},
address = {Anchorage, Alaska, USA},
month = {May}
} abstract: We present an interface for controlling mobile robots that combines aspects of graphical trajectory specification and state-based programming. This work is motivated by common tasks executed by our underwater vehicles, although we illustrate a mode of interaction that is applicable to mobile robotics in general. The key aspect of our approach is to provide an intuitive linkage between the graphical visualization of regions of interest in the environment, and activities relevant to these regions. In addition to introducing this novel programming@inproceedings{Girdhar2010,
author = {Yogesh Girdhar and Gregory Dudek},
title = {ONSUM: A System for Generating Online Navigation Summaries},
booktitle = {Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS2010)},
year = {2010},
address = {Taipei, Taiwan},
month = {October},
note = {Nominee for ICROS best application paper award from 972 accepted papers}
} abstract: We propose an algorithm for generating navigation summaries. Navigation summaries are a specialization of video summaries, where the focus is on video collected by a mobile robot, on a specified trajectory. We are interested in finding a few images that epitomize the visual experience of a robot as it traverses a terrain. This paper presents a novel approach to generating summaries in form of a set of images, where the decision to include the image in the summary set is made online. Our focus is on the case where the number of observations@article{marinakis2010pure,
abstract = {In this paper, we investigate a pure form of the topological mapping problem in mobile robotics. We consider the mapping ability of a robot navigating a graph-like world in which it is able to assign a relative ordering to the edges, leaving a vertex with reference to the edge by which it arrived but is unable to associate a unique label with any vertex or edge. Our work extends and builds upon earlier approaches in this problem domain, which are based on construction of exploration tree of plausible world models. The main contributions of the},
author = {Marinakis, Dimitri and Dudek, Gregory},
journal = {IEEE Transactions on Robotics},
number = {6},
pages = {1051--1064},
pub_year = {2010},
publisher = {IEEE},
title = {Pure topological mapping in mobile robotics},
venue = {IEEE Transactions on Robotics},
volume = {26}
} abstract: In this paper, we investigate a pure form of the topological mapping problem in mobile robotics. We consider the mapping ability of a robot navigating a graph-like world in which it is able to assign a relative ordering to the edges, leaving a vertex with reference to the edge by which it arrived but is unable to associate a unique label with any vertex or edge. Our work extends and builds upon earlier approaches in this problem domain, which are based on construction of exploration tree of plausible world models. The main contributions of the@inproceedings{sattar2010reducing,
title={Reducing Uncertainty in Human-Robot Interaction: A Cost Analysis Approach},
author={Sattar, Junaed and Dudek, Gregory},
booktitle={Proceedings of the Twelfth International Symposium on Experimental Robotics (ISER 2010)},
year={2010},
address={New Delhi and Agra, India},
month={December}
} abstract: We present a technique for robust human-robot interaction taking into consideration uncertainty in input and task execution costs incurred by the robot. Specifically, this research aims to quantitatively model confirmation feedback, as required by a robot while communicating with a human operator to perform a particular task. Our goal is to model human-robot interaction from the perspective of risk minimization, taking into account errors in communication,“risk” involved in performing the required task, and task execution costs@inproceedings{Rekleitis2010,
author = {Ioannis Rekleitis and Yasmina Schoueri and Philippe Gigu\`ere and Junaed Sattar and Gregory Dudek},
title = {Telepresence Across the Ocean},
booktitle = {Proceedings of the Seventh Canadian Conference on Computer and Robot Vision (CRV 2010)},
year = {2010},
pages = {261--268},
address = {Ottawa, Ontario, Canada},
month = {May}
} abstract: We describe the development and deployment of a system for long-distance remote observation of robotic operations. The system we have developed is targeted to exploration, multi-participant interaction, and tele-learning. In particular, we used this system with a robot deployed in an underwater environment in order to produce interactive web-casts of scientific material. The system used a combination of robotic and networking technologies and was deployed and evaluated in a context where students in a classroom were able to@inproceedings{Girdhar2011,
author = {Yogesh Girdhar and Gregory Dudek},
title = {Offline Navigation Summaries},
booktitle = {Proceedings of the IEEE International Conference on Robotics and Automation (ICRA2011)},
year = {2011},
pages = {5769--5775},
address = {Shanghai, China},
month = {May 9--13}
} abstract: Our objective is to find a small set of images that summarize a robot's visual experience along a path. We present a novel on-line algorithm for this task. This algorithm is based on a new extension to the classical Secretaries Problem. We also present an extension to the idea of Bayesian Surprise, which we then use to measure the fitness of an image as a summary image.@book{dudek2024computational,
title={Computational Principles of Mobile Robotics (2nd edition)},
author={Dudek, Gregory and Jenkin, Michael},
edition={3rd},
year={2010},
publisher={Cambridge University Press},
pages={450}
} abstract: Computational Principles of Mobile Robotics 2nd edition@inproceedings{sattar2009vision,
title={A Vision-based Control and Interaction Framework for a Legged Underwater Robot},
author={Sattar, Junaed and Dudek, Gregory},
booktitle={Proc. Sixth Canadian Conference on Robot Vision (CRV)},
pages={329--336},
year={2009},
address={Kelowna, BC, Canada},
note={Award for best robotics paper}
} abstract: We present a vision-based control and interaction framework for mobile robots, and describe its implementation in a legged amphibious robot. The control scheme enables the robot to navigate, follow targets of interest, and interact with human operators. The visual framework presented in this paper enables deployment of the vehicle in underwater environments along with a human scuba diver as the operator, without requiring any external tethered control. We present the current implementation of this framework in our particular family of@article{giguere2009clustering,
abstract = {In this paper we are interested in autonomous vehicles that can automatically develop terrain classifiers without human interaction or feedback. A key issue is the clustering of time-series data collected by the sensors of a ground-based vehicle moving over several terrain surfaces (eg concrete or soil). In this context, we present a novel off-line windowless clustering algorithm that exploits time-dependency between samples. In terrain coverage, sets of sensory measurements are returned that are spatially, and hence temporally},
author = {Giguere, Philippe and Dudek, Gregory},
journal = {Autonomous Robots},
pages = {171--186},
pub_year = {2009},
publisher = {Springer},
title = {Clustering sensor data for autonomous terrain identification using time-dependency},
venue = {Autonomous Robots},
volume = {26}
} abstract: In this paper we are interested in autonomous vehicles that can automatically develop terrain classifiers without human interaction or feedback. A key issue is the clustering of time-series data collected by the sensors of a ground-based vehicle moving over several terrain surfaces (eg concrete or soil). In this context, we present a novel off-line windowless clustering algorithm that exploits time-dependency between samples. In terrain coverage, sets of sensory measurements are returned that are spatially, and hence temporally@inproceedings{pomerantz2009context,
title={Context Dependent Movie Recommendations Using a Hierarchical Bayesian Model},
author={Pomerantz, Daniel and Dudek, Gregory},
booktitle={Proceedings of the 22nd Canadian Conference on Artificial Intelligence, Canadian AI 2009},
year={2009},
location={Kelowna, British Columbia, Canada}
} abstract: We use a hierarchical Bayesian approach to model user preferences in different contexts or settings. Unlike many previous recommenders, our approach is content-based. We assume that for each context, a user has a different set of preference weights which are linked by a common,“generic context” set of weights. The approach uses Expectation Maximization (EM) to estimate both the generic context weights and the context specific weights. This improves upon many current recommender systems that do not incorporate context into the@article{mills2009image,
abstract = {This paper presents a new combination of techniques to create pleasing and physically consistent image mosaics despite the presence of moving objects in the scene. The technique uses heuristic seam selection in the intensity and gradient-domains to choose which pixels to use from each image and then blends them smoothly to create the final mosaic. We demonstrate illustrative results obtained by comparing and contrasting our output with that obtained from four representative existing image mosaic systems. One of},
author = {Mills, Alec and Dudek, Gregory},
journal = {Image and Vision Computing},
number = {10},
pages = {1593--1602},
pub_year = {2009},
publisher = {Elsevier},
title = {Image stitching with dynamic elements},
venue = {Image and Vision Computing},
volume = {27}
} abstract: This paper presents a new combination of techniques to create pleasing and physically consistent image mosaics despite the presence of moving objects in the scene. The technique uses heuristic seam selection in the intensity and gradient-domains to choose which pixels to use from each image and then blends them smoothly to create the final mosaic. We demonstrate illustrative results obtained by comparing and contrasting our output with that obtained from four representative existing image mosaic systems. One of@inproceedings{meger2009inferring,
title={Inferring a Probability Distribution Function for the Pose of a Sensor Network using a Mobile Robot},
author={Meger, David and Marinakis, Dimitri and Rekleitis, Ioannis and Dudek, Gregory},
booktitle={Proc. IEEE International Conference on Robotics and Automation (ICRA2009)},
pages={Kobe, Japan},
year={2009},
month={May 12--17}
} abstract: In this paper we present an approach for localizing a sensor network augmented with a mobile robot which is capable of providing inter-sensor pose estimates through its odometry measurements. We present a stochastic algorithm that samples efficiently from the probability distribution for the pose of the sensor network by employing Rao-Blackwellization and a proposal scheme which exploits the sequential nature of odometry measurements. Our algorithm automatically tunes itself to the problem instance and includes a principled@inproceedings{Girdhar2009,
author = {Yogesh Girdhar and Gregory Dudek},
title = {Optimal Online Data Sampling or How to Hire the Best Secretaries},
booktitle = {Proceedings of the Sixth Canadian Conference on Computer and Robot Vision},
year = {2009},
address = {Kelowna, British Columbia, Canada},
month = {May}
} abstract: The problem of online sampling of data, can be seen as a generalization of the classical secretary problem. The goal is to maximize the probability of picking the k highest scoring samples in our data, making the decision to select or reject a sample online. We present a new and simple online algorithm to optimally make this selection. We then apply this algorithm to a sequence of images taken by a mobile robot, with the goal of identifying the most interesting and informative images.@inproceedings{sattar2009robust,
title={Robust Servo-control for Underwater Robots using Banks of Visual Filters},
author={Sattar, Junaed and Dudek, Gregory},
booktitle={Proc. IEEE International Conference on Robotics and Automation (ICRA2009)},
pages={3583--3588},
year={2009},
address={Kobe, Japan},
month={May 12--17}
} abstract: We present an application of machine learning to the semi-automatic synthesis of robust servo-trackers for underwater robotics. In particular, we investigate an approach based on the use of Boosting for robust visual tracking of color objects in an underwater environment. To this end, we use AdaBoost, the most common variant of the Boosting algorithm, to select a number of low-complexity but moderately accurate color feature trackers and we combine their outputs. The novelty of our approach lies in the design of this family of weak trackers@article{marinakis2009self,
abstract = {When a network of vision-based sensors is emplaced in an environment for applications such as surveillance or monitoring the spatial relationships between the sensing units must be inferred or computed for self-calibration purposes. In this paper we describe a technique to solve one aspect of this self-calibration problem: automatically determining the topology and connectivity information of a network of cameras based on a statistical analysis of observed motion in the environment. While the technique can use labels from reliable},
author = {Marinakis, Dimitri and Dudek, Gregory},
journal = {Image and vision computing},
number = {1-2},
pages = {116--130},
pub_year = {2009},
publisher = {Elsevier},
title = {Self-calibration of a vision-based sensor network},
venue = {Image and vision computing},
volume = {27}
} abstract: When a network of vision-based sensors is emplaced in an environment for applications such as surveillance or monitoring the spatial relationships between the sensing units must be inferred or computed for self-calibration purposes. In this paper we describe a technique to solve one aspect of this self-calibration problem: automatically determining the topology and connectivity information of a network of cameras based on a statistical analysis of observed motion in the environment. While the technique can use labels from reliable@inproceedings{dudek2008sensor,
abstract = {In this paper, we present behaviors and interaction modes for a small underwater robot. In particular, we address some challenging issues arising from the underwater environment: visual processing, interactive communication with an underwater crew, and finally orientation and motion of the vehicle through a hovering mode. The visual processing consist of target tracking using various techniques (color blob, color histogram and mean shift). The underwater communication is achieved through printed cards with virtual markers},
author = {Dudek, Gregory and Giguere, Philippe and Sattar, Junaed},
booktitle = {Experimental Robotics: The 10th International Symposium on Experimental Robotics},
organization = {Springer},
pages = {267--276},
pub_year = {2008},
title = {Sensor-based behavior control for an autonomous underwater vehicle},
venue = {Experimental Robotics: The 10th …}
} abstract: In this paper, we present behaviors and interaction modes for a small underwater robot. In particular, we address some challenging issues arising from the underwater environment: visual processing, interactive communication with an underwater crew, and finally orientation and motion of the vehicle through a hovering mode. The visual processing consist of target tracking using various techniques (color blob, color histogram and mean shift). The underwater communication is achieved through printed cards with virtual markers@inproceedings{giguere2009surface,
author = {Philippe Gigu\`ere and Gregory Dudek},
title = {Surface Identification Using Simple Contact Dynamics for Mobile Robots},
booktitle = {Proc. IEEE International Conference on Robotics and Automation (ICRA2009)},
year = {2009},
address = {Kobe, Japan},
month = {May 12-17}
} abstract: This paper describes an approach to surface identification in the context of mobile robotics, applicable to supervised and unsupervised learning. The identification is based on analyzing the tip acceleration patterns induced in a metallic rod, dragged along a surface that is to be identified. Eight features in time and frequency domains are used for classification. Results show that for ten type of indoor and outdoor surfaces, reliable identification can be achieved (90.0 and 94.6 percent for a 1 and 4 seconds time-window@inproceedings{Dudek2009,
author = {G. Dudek and J.-P. Lobos},
title = {Towards Navigation Summaries: Automated Production of a Synopsis of a Robot Trajectories},
booktitle = {Proc. Conference on Computer and Robot Vision},
year = {2009},
pages = {93--100},
doi = {10.1109/CRV.2009.43},
address = {Kelowna, BC, Canada}
} abstract: In this paper we describe an approach to computing a navigation summary: a visual synopsis of the notable images that characterize a trajectory. We use a combination of PCA and supplementary features to ensure both converge of the trajectory and appearance spaces. The results obtained from a series of experiments are promising and provide us with a method to index and classify video footage from our robot.@inproceedings{sattar2009underwater,
title={Underwater Human-Robot Interaction via Biological Motion Identification},
author={Sattar, Junaed and Dudek, Gregory},
booktitle={Proc. Robotics: Science and Systems V},
year={2009},
address={Seattle, WA, USA},
month={June-July}
} abstract: We present an algorithm for underwater robots to visually detect and track human motion. Our objective is to enable human-robot interaction by allowing a robot to follow behind a human moving in (up to) six degrees of freedom. In particular, we have developed a system to allow a robot to detect, track and follow a scuba diver by using frequencydomain detection of biological motion patterns. The motion of biological entities is characterized by combinations of periodic motions which are inherently distinctive. This is especially true of@inproceedings{Giguere2009,
author = {Philippe Gigu\`ere and Christopher Prahacs and Nicolas Plamondon and Katrine Turgeon and Gregory Dudek},
title = {Unsupervised Learning of Terrain Appearance for Automated Coral Reef Exploration},
booktitle = {Proceedings of the Sixth Canadian Conference on Computer and Robot Vision},
year = {2009},
address = {Kelowna, British Columbia, Canada},
month = {May}
} abstract: We describe a navigation and coverage system based on unsupervised learning driven by visual input. Our objective is to allow a robot to remain continuously moving above a terrain of interest using visual feedback to avoid leaving this region. As a particular application domain, we are interested in doing this in open water, but the approach makes few domain-specific assumptions. Specifically, our system employed an unsupervised learning technique to train a k-Nearest Neighbor classifier to distinguish between images of different@article{chen2009auto,
title={Auto-correlation wavelet support vector machine},
author={Chen, Guangyi and Dudek, Gregory},
journal={Image Vision Computing},
volume={27},
number={8},
pages={1040--1046},
year={2009},
publisher={Elsevier},
doi={10.1016/j.imavis.2008.09.006}
} abstract: A support vector machine (SVM) with the auto-correlation of a compactly supported wavelet as a kernel is proposed in this paper. The authors prove that this kernel is an admissible support vector kernel. The main advantage of the auto-correlation of a compactly supported wavelet is that it satisfies the translation invariance property, which is very important for its use in signal processing. Also, we can choose a better wavelet by selecting from different wavelet families for our auto-correlation wavelet kernel. This is because for different@inproceedings{sattar2008boosting,
author = {Junaed Sattar and Gregory Dudek},
title = {A Boosting Approach to Visual Servo-Control of an Underwater Robot},
booktitle = {Proceedings of the 11th International Symposium on Experimental Robotics, ISER},
year = {2008},
address = {Athens, Greece},
month = {July}
} abstract: We present an application of the ensemble learning algorithm in the area of visual tracking and servoing. In particular, we investigate an approach based on the Boosting technique for robust visual tracking of color objects in an underwater environment. To this end, we use AdaBoost, the most common variant of the Boosting algorithm, to select a number of low-complexity but moderately accurate color feature trackers and we combine their outputs. From a significantly large number of “weak” color trackers, the training process selects those@inproceedings{xu2008natural,
title={A Natural Gesture Interface for Operating Robotic Systems},
author={Xu, Anqi and Sattar, Junaed and Dudek, Gregory},
booktitle={Proceedings of the IEEE International Conference on Robotics and Automation (ICRA)},
pages={3557--3563},
year={2008},
location={Pasadena, California, USA}
} abstract: A gesture-based interaction framework is presented for controlling mobile robots. This natural interaction paradigm has few physical requirements, and thus can be deployed in many restrictive and challenging environments. We present an implementation of this scheme in the control of an underwater robot by an on-site human operator. The operator performs discrete gestures using engineered visual targets, which are interpreted by the robot as parametrized actionable commands. By combining the symbolic alphabets resulting@article{chen2009auto,
title={Auto-correlation wavelet support vector machine},
author={Chen, Guangyi and Dudek, Gregory},
journal={Image Vision Computing},
volume={27},
number={8},
pages={1040--1046},
year={2009},
publisher={Elsevier},
doi={10.1016/j.imavis.2008.09.006}
} abstract: A support vector machine (SVM) with the auto-correlation of a compactly supported wavelet as a kernel is proposed in this paper. The authors prove that this kernel is an admissible support vector kernel. The main advantage of the auto-correlation of a compactly supported wavelet is that it satisfies the translation invariance property, which is very important for its use in signal processing. Also, we can choose a better wavelet by selecting from different wavelet families for our auto-correlation wavelet kernel. This is because for different@inproceedings{giguere2008clustering,
title={Clustering Sensor Data for Terrain Identification using a Windowless Algorithm},
author={Gigu{\`e}re, Philippe and Dudek, Gregory},
booktitle={Proc. Robotics Science and System (RSS)},
year={2008},
month={June},
address={Zurich, Switzerland}
} abstract: in autonomous systems that can automatically develop terrain A key issue is clustering of sensor data from the same terrain on data collected from a mobile robot enables robust terrain@inproceedings{sattar2008enabling,
title={Enabling Autonomous Capabilities in Underwater Robotics},
author={Sattar, Junaed and Chiu, Olivia and Rekleitis, Ioannis and Gigu{\`e}re, Philippe and Mills, Alec and Plamondon, Nicolas and Prahacs, Chris and Girdhar, Yogesh and Nahon, Meyer and Lobos, John-Paul and Dudek, Gregory},
booktitle={Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
pages={pp. 3628--3634},
year={2008},
month={September},
address={Nice, France}
} abstract: Underwater operations present unique challenges and opportunities for robotic applications. These can be attributed in part to limited sensing capabilities, and to locomotion behaviours requiring control schemes adapted to specific tasks or changes in the environment. From enhancing teleoperation procedures, to providing high-level instruction, all the way to fully autonomous operations, enabling autonomous capabilities is fundamental for the successful deployment of underwater robots. This paper presents an overview of the approaches used@article{torres2008inter,
abstract = {In this article we present a method for automatically recovering complete and dense depth maps of an indoor environment by fusing incomplete data for the 3D environment modeling problem. The geometry of indoor environments is usually extracted by acquiring a huge amount of range data and registering it. By acquiring a small set of intensity images and a very limited amount of range data, the acquisition process is considerably simplified, saving time and energy consumption. In our method, the intensity and partial range data are},
author = {Torres-M{\'e}ndez, Luz A and Dudek, Gregory},
journal = {International journal of computer vision},
pages = {137--158},
pub_year = {2008},
publisher = {Springer},
title = {Inter-image statistics for 3d environment modeling},
venue = {International journal of computer vision},
volume = {79}
} abstract: In this article we present a method for automatically recovering complete and dense depth maps of an indoor environment by fusing incomplete data for the 3D environment modeling problem. The geometry of indoor environments is usually extracted by acquiring a huge amount of range data and registering it. By acquiring a small set of intensity images and a very limited amount of range data, the acquisition process is considerably simplified, saving time and energy consumption. In our method, the intensity and partial range data are@article{marinakis2008occam,
abstract = {We present a method for inferring the topology of a sensor network given nondiscriminating observations of activity in the monitored region. This is accomplished based on no prior knowledge of the relative locations of the sensors and weak assumptions regarding environmental conditions. Our approach employs a two-level reasoning system made up of a stochastic expectation maximization algorithm and a higher level search strategy employing the principle of Occam's Razor to look for the simplest solution explaining the},
author = {Marinakis, Dimitri and Dudek, Gregory},
journal = {IEEE Transactions on Robotics},
number = {2},
pages = {293--306},
pub_year = {2008},
publisher = {IEEE},
title = {Occam's Razor applied to network topology inference},
venue = {IEEE Transactions on Robotics},
volume = {24}
} abstract: We present a method for inferring the topology of a sensor network given nondiscriminating observations of activity in the monitored region. This is accomplished based on no prior knowledge of the relative locations of the sensors and weak assumptions regarding environmental conditions. Our approach employs a two-level reasoning system made up of a stochastic expectation maximization algorithm and a higher level search strategy employing the principle of Occam's Razor to look for the simplest solution explaining the@book{dudek2008digital,
author = {Gregory Dudek},
title = {Digital Television at Home: Satellite, Cable and Over-The-Air},
year = {2008},
publisher = {Y1D Books},
isbn = {978-9809915-0-5}
} abstract: Digital Television at Home: Satellite, Cable and Over-The-Air@inproceedings{sattar2007visual,
author = {Junaed Sattar and Anqi Xu and Gregory Dudek},
title = {A Visual Language for Robot Control and Programming: A Human-Interface Study},
booktitle = {Proceedings of the IEEE International Conference of Robotics and Automation (ICRA)},
year = {2007},
address = {Rome, Italy},
month = {April},
pages = {2507--2513}
} abstract: We describe an interaction paradigm for controlling a robot using hand gestures. In particular, we are interested in the control of an underwater robot by an on-site human operator. Under this context, vision-based control is very attractive, and we propose a robot control and programming mechanism based on visual symbols. A human operator presents engineered visual targets to the robotic system, which recognizes and interprets them. This paper describes the approach and proposes a specific gesture language called" RoboChat"@article{dudek2007aqua,
abstract = {AQUA, an amphibious robot that swims via the motion of its legs rather than using thrusters and control surfaces for propulsion, can walk along the shore, swim along the surface in open water, or walk on the bottom of the ocean. The vehicle uses a variety of sensors to estimate its position with respect to local visual features and provide a global frame of reference},
author = {Dudek, Gregory and Giguere, Philippe and Prahacs, Chris and Saunderson, Shane and Sattar, Junaed and Torres-Mendez, Luz-Abril and Jenkin, Michael and German, Andrew and Hogue, Andrew and Ripsman, Arlene and others},
journal = {Computer},
number = {1},
pages = {46--53},
pub_year = {2007},
publisher = {IEEE},
title = {Aqua: An amphibious autonomous robot},
venue = {Computer},
volume = {40}
} abstract: AQUA, an amphibious robot that swims via the motion of its legs rather than using thrusters and control surfaces for propulsion, can walk along the shore, swim along the surface in open water, or walk on the bottom of the ocean. The vehicle uses a variety of sensors to estimate its position with respect to local visual features and provide a global frame of reference@inproceedings{marinakis2007hybrid,
title={Hybrid Inference for Sensor Network Localization using a Mobile Robot},
author={Marinakis, Dimitri and Meger, David and Rekleitis, Ioannis and Dudek, Gregory},
booktitle={Proceedings of the National Conference on Artificial Intelligence (AAAI)},
year={2007},
address={Vancouver, Canada},
month={July}
} abstract: In this paper, we consider a hybrid solution to the sensor network position inference problem, which combines a real-time filtering system with information from a more expensive, global inference procedure to improve accuracy and prevent divergence. Many online solutions for this problem make use of simplifying assumptions, such as Gaussian noise models and linear system behaviour and also adopt a filtering strategy which may not use available information optimally. These assumptions allow near real-time inference@inproceedings{marinakis2007learning,
title={Learning Network Topology from Simple Sensor Data},
author={Marinakis, Dimitri and Gigu{\`e}re, Philippe and Dudek, Gregory},
booktitle={20th Canadian Conference on Artificial Intelligence},
year={2007},
address={Montreal, Canada}
} abstract: In this paper, we present an approach for recovering a topological map of the environment using only detection events from a deployed sensor network. Unlike other solutions to this problem, our technique operates on timestamp free observational data; ie no timing information is exploited by our algorithm except the ordering. We first give a theoretical analysis of this version of the problem, and then we show that by considering a sliding window over the observations, the problem can be re-formulated as a version of set@article{rao2007randomized,
abstract = {The problem of minimum distance localization in environments that may contain self-similarities is addressed. A mobile robot is placed at an unknown location inside a 2 D self-similar polygonal environment P. The robot has a map of P and can compute visibility data through sensing. However, the self-similarities in the environment mean that the same visibility data may correspond to several different locations. The goal, therefore, is to determine the robot's true initial location while minimizing the distance traveled by the robot},
author = {Rao, Malvika and Dudek, Gregory and Whitesides, Sue},
journal = {The International Journal of Robotics Research},
number = {9},
pages = {917--933},
pub_year = {2007},
publisher = {Sage Publications Sage UK: London, England},
title = {Randomized algorithms for minimum distance localization},
venue = {The International Journal of …},
volume = {26}
} abstract: The problem of minimum distance localization in environments that may contain self-similarities is addressed. A mobile robot is placed at an unknown location inside a 2 D self-similar polygonal environment P. The robot has a map of P and can compute visibility data through sensing. However, the self-similarities in the environment mean that the same visibility data may correspond to several different locations. The goal, therefore, is to determine the robot's true initial location while minimizing the distance traveled by the robot@inproceedings{marinakis2007topological,
title={Topological Mapping through Distributed, Passive Sensors},
author={Marinakis, Dimitri and Dudek, Gregory},
booktitle={Proceedings of the International Joint Conference on Artificial Intelligence (IJCAI07)},
pages={2147--2152},
year={2007},
address={Hyderabad, India}
} abstract: In this paper we address the problem of inferring the topology, or inter-node navigability, of a sensor network given non-discriminating observations of activity in the environment. By exploiting motion present in the environment, our approach is able to recover a probabilistic model of the sensor network connectivity graph and the underlying traffic trends. We employ a reasoning system made up of a stochastic Expectation Maximization algorithm and a higher level search strategy employing the principle of Occam's Razor to look for the@inproceedings{marinakis2007topological,
title={Topological Mapping with Weak Sensory Data},
author={Marinakis, Dimitri and Dudek, Gregory},
booktitle={Proceedings of the National Conference on Artificial Intelligence (AAAI)},
year={2007},
address={Vancouver, Canada},
month={July}
} abstract: In this paper, we consider the exploration of topological environments by a robot with weak sensory capabilities. We assume only that the robot can recognize when it has reached a vertex, and can assign a cyclic ordering to the edges leaving a vertex with reference to the edge it arrived from. Given this limited sensing capability, and without the use of any markers or additional information, we will show that the construction of a topological map is still feasible. This is accomplished through both the exploration strategy which is designed to@inproceedings{sattar2007dive,
title={Where is your dive buddy: Tracking scuba divers using spatio-temporal features},
author={Sattar, Junaed and Dudek, Gregory},
booktitle={Proc. of the IEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
pages={3654--3659},
year={2007},
address={San Diego, California, USA},
month={October-November}
} abstract: Where is your dive buddy: Tracking scuba divers using spatio-temporal features@inproceedings{marinakis2006practical,
title={A Practical Algorithm for Network Topology Inference},
author={Marinakis, Dimitri and Dudek, Gregory},
booktitle={Proceedings of the IEEE International Conference on Robotics and Automation (ICRA06)},
pages={3108--3115},
year={2006},
organization={IEEE},
address={Orlando, Florida},
month={May}
} abstract: When a network of robots or static sensors is emplaced in an environment, the spatial relationships between the sensing units must be inferred or computed for most key applications. In this paper we present a Monte Carlo expectation maximization algorithm for recovering the connectivity information (ie topological map) of a network using only detection events from deployed sensors. The technique is based on stochastically reconstructing samples of plausible agent trajectories allowing for the possibility of@inproceedings{meger2006autonomous,
title={Autonomous Mobile Robot Mapping of a Camera Sensor Network},
author={Meger, David and Rekleitis, Ioannis and Dudek, Gregory},
booktitle={Proc. 8th International Symposium on Distributed Autonomous Robotic Systems (DARS)},
year={2006}
} abstract: @inproceedings{giguere2006characterization,
title={Characterization and Modeling of Rotational Responses for an Oscillating Foil Underwater Robot},
author={Gigu{\`e}re, Philippe and Prahacs, Chris and Dudek, Gregory},
booktitle={Proc. IEEE/RSJ/GI International Conference on Intelligent Robots and Systems (IROS)},
pages={October},
year={2006},
address={Beijing, China}
} abstract: In order to better understand the behavior of the underwater robot developed at our laboratory, a simple but relatively good model of the underwater behavior of the robot had to be developed. In order to be useful for model-based control techniques onboard the robot, the model had to have low computing requirements, yet be complex enough to capture the transient response of the robot. To achieve this, a system identification approach was taken by first capturing the robot response to various inputs, and then matching them to a simple@inproceedings{giguere2006environment,
title={Environment Identification for a Running Robot Using Inertial and Actuator Cues},
author={Gigu{\`e}re, Philippe and Prahacs, Chris and Saunderson, Shane and Dudek, Gregory},
booktitle={Proc. Robotics Science and System (RSS)},
year={2006},
month={August},
address={Philadelphia, U.S.A}
} abstract: Environment Identification for a Running Robot Using Inertial and Actuator Cues@inproceedings{sattar2006fourier,
author = {Junaed Sattar and Eric Bourque and Philippe Gigu\`ere and Gregory Dudek},
title = {Fourier tags: Smoothly degradable fiducial markers for use in human-robot interaction},
booktitle = {Proceedings of the Canadian Conference on Computer and Robot Vision (CRV06)},
year = {2006},
pages = {22--29},
address = {Quebec City, Quebec}
} abstract: In this paper we introduce the Fourier tag, a synthetic fiducial marker used to visually encode information and provide controllable positioning. The Fourier tag is a synthetic target akin to a bar-code that specifies multi-bit information which can be efficiently and robustly detected in an image. Moreover, the Fourier tag has the beneficial property that the bit string it encodes has variable length as a function of the distance between the camera and the target. This follows from the fact that the effective resolution decreases as an effect of@article{theberge2006gone,
abstract = {This paper describes a mechanical hexapod designed by a research group from McGill University. Called Aqua, it is latest in a series of seagoing robots. With six rotating flippers, this machine is amphibious, capable of both walking and swimming. The group aims to develop an underwater vehicle that can autonomously explore and collect data in aquatic environments while surviving the harsh saltwater conditions and often turbulent waters of the open sea},
author = {Theberge, M and Dudek, G},
journal = {IEEE spectrum},
number = {6},
pages = {38--43},
pub_year = {2006},
publisher = {IEEE},
title = {Gone swimmin'[seagoing robots]},
venue = {IEEE spectrum},
volume = {43}
} abstract: This paper describes a mechanical hexapod designed by a research group from McGill University. Called Aqua, it is latest in a series of seagoing robots. With six rotating flippers, this machine is amphibious, capable of both walking and swimming. The group aims to develop an underwater vehicle that can autonomously explore and collect data in aquatic environments while surviving the harsh saltwater conditions and often turbulent waters of the open sea@inproceedings{garden2006mixed,
title={Mixed Collaborative and Content-Based Filtering with User-Contributed Semantic Features},
author={Garden, Mat and Dudek, Gregory},
booktitle={Proceedings of the National Conference on Artificial Intelligence (AAAI)},
year={2006},
address={Boston, MA},
month={July}
} abstract: We describe a recommender system which uses a unique combination of content-based and collaborative methods to suggest items of interest to users, and also to learn and exploit item semantics. Recommender systems typically use techniques from collaborative filtering, in which proximity measures between users are formulated to generate recommendations, or content-based filtering, in which users are compared directly to items. Our approach uses similarity measures between users, but also directly measures the attributes of items that@inproceedings{sattar2006performance,
title={On the Performance of Color Tracking Algorithms for Underwater Robots under Varying Lighting and Visibility},
author={Sattar, Junaed and Dudek, Gregory},
booktitle={Proceedings of the IEEE International Conference on Robotics and Automation (ICRA06)},
pages={3550--3555},
year={2006},
address={Orlando, Florida}
} abstract: We consider the use of visual target tracking for autonomous steering of an underwater robot. In this context, we consider a performance comparison for three key visual tracking algorithms used for servo control. We present a comparative study of the performance in underwater environments of three tracking algorithms that are widely used in vision applications. Variations in illumination, suspended particles and a resulting reduction in visibility hinders vision systems from performing satisfactorily in marine environments; at@inproceedings{Marinakis2006,
author = {Dimitri Marinakis and Gregory Dudek},
title = {Probabilistic Self-Localization for Sensor Networks},
booktitle = {Proc. of the AAAI National Conference on Artificial Intelligence},
year = {2006},
month = {July},
address = {Boston, Massachusetts, USA}
} abstract: This paper describes a technique for the probabilistic self-localization of a sensor network based on noisy inter-sensor range data. Our method is based on a number of parallel instances of Markov Chain Monte Carlo (MCMC). By combining estimates drawn from these parallel chains, we build up a representation of the underlying probability distribution function (PDF) for the network pose. Our approach includes sensor data incrementally in order to avoid local minima and is shown to produce meaningful results efficiently. We return@inproceedings{burfoot2006rrt,
title={RRT-Plan: a Randomized Algorithm for STRIPs planning},
author={Burfoot, Daniel and Pineau, Joelle and Dudek, Gregory},
booktitle={Proc. of the International Conference on Automated Planning and Scheduling (ICAPS06)},
year={2006},
address={Cumbria, UK},
month={June}
} abstract: We propose a randomized STRIPS planning algorithm called RRT-Plan. This planner is inspired by the idea of Rapidly exploring Random Trees, a concept originally designed for use in continuous path planning problems. Issues that arise in the conversion of RRTs from continuous to discrete spaces are discussed, and several additional mechanisms are proposed to improve performance. Our experimental results indicate that RRT-Plan is competitive with the state of the art in STRIPS planning.@inproceedings{dudek2008sensor,
abstract = {In this paper, we present behaviors and interaction modes for a small underwater robot. In particular, we address some challenging issues arising from the underwater environment: visual processing, interactive communication with an underwater crew, and finally orientation and motion of the vehicle through a hovering mode. The visual processing consist of target tracking using various techniques (color blob, color histogram and mean shift). The underwater communication is achieved through printed cards with virtual markers},
author = {Dudek, Gregory and Giguere, Philippe and Sattar, Junaed},
booktitle = {Experimental Robotics: The 10th International Symposium on Experimental Robotics},
organization = {Springer},
pages = {267--276},
pub_year = {2008},
title = {Sensor-based behavior control for an autonomous underwater vehicle},
venue = {Experimental Robotics: The 10th …}
} abstract: In this paper, we present behaviors and interaction modes for a small underwater robot. In particular, we address some challenging issues arising from the underwater environment: visual processing, interactive communication with an underwater crew, and finally orientation and motion of the vehicle through a hovering mode. The visual processing consist of target tracking using various techniques (color blob, color histogram and mean shift). The underwater communication is achieved through printed cards with virtual markers@article{rekleitis2006simultaneous,
abstract = {In this paper we examine issues of localization, exploration, and planning in the context of a hybrid robot/camera-network system. We exploit the ubiquity of camera networks to use them as a source of localization data. Since the Cartesian position of the cameras in most networks is not known accurately, we consider the issue of how to localize such cameras. To solve this hybrid localization problem, we divide it into a local problem of camera-parameter estimation combined with a global planning and navigation problem. We solve the local},
author = {Rekleitis, Ioannis and Meger, David and Dudek, Gregory},
journal = {Robotics and Autonomous Systems},
number = {11},
pages = {921--932},
pub_year = {2006},
publisher = {Elsevier},
title = {Simultaneous planning, localization, and mapping in a camera sensor network},
venue = {Robotics and Autonomous Systems},
volume = {54}
} abstract: In this paper we examine issues of localization, exploration, and planning in the context of a hybrid robot/camera-network system. We exploit the ubiquity of camera networks to use them as a source of localization data. Since the Cartesian position of the cameras in most networks is not known accurately, we consider the issue of how to localize such cameras. To solve this hybrid localization problem, we divide it into a local problem of camera-parameter estimation combined with a global planning and navigation problem. We solve the local@inproceedings{Torres-Mendez2006,
author = {Luz Abril Torres-Mendez and Gregory Dudek},
title = {Statistics of Visual and Partial Depth Data for Mobile Robot Environment Modeling},
booktitle = {Proc. Mexican International Conference on Artificial Intelligence (MICAI)},
year = {2006},
month = {November}
} abstract: In mobile robotics, the inference of the 3D layout of large-scale indoor environments is a critical problem for achieving exploration and navigation tasks. This article presents a framework for building a 3D model of an indoor environment from partial data using a mobile robot. The modeling of a large-scale environment involves the acquisition of a huge amount of range data to extract the geometry of the scene. This task is physically demanding and time consuming for many real systems. Our approach overcomes this problem by allowing a@inproceedings{johns2006urban,
author = {Derek Johns and Gregory Dudek},
title = {Urban Position Estimation from One Dimensional Visual Cues},
booktitle = {Proceedings of the Canadian Conference on Computer and Robot Vision (CRV06)},
year = {2006},
pages = {22--29},
address = {Quebec City, Quebec},
month = {June}
} abstract: We consider the problem of vision-based position estimation in urban environments. In particular, we are interested in position estimation from visual cues, but using only limited computational resources. Our particular solution to this problem is based on representing the variability of the" horizon" of the cityscape when seen from within the city; that is, the outlines of the rooftops of adjacent buildings. By encoding the image using only such a one-dimensional contour, we obtain an image encoding that is exceedingly compact. This, in@inproceedings{sattar2005visual,
title={A Visual Servoing System for an Aquatic Swimming Robot},
author={Sattar, Junaed and Gigu{\`e}re, Philippe and Prahacs, Chris and Dudek, Gregory},
booktitle={Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
pages={1483--1488},
year={2005},
address={Edmonton, Alberta, Canada},
month={August}
} abstract: A Visual Servoing System for an Aquatic Swimming Robot@inproceedings{jenkin2005visually,
author = {Michael Jenkin and Chris Prahacs and Andrew Hogue and Junaed Sattar and Philippe Gigu\`ere and Andrew German and Hui Liu and Shane Saunderson and Arlene Ripsman and Saul Simhon and Luz Abril Torres-Mendez and Evangelos Milios and Pifu Zhang and Ioannis Rekleitis and Gregory Dudek},
title = {A Visually Guided Swimming Robot},
booktitle = {Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
year = {2005},
pages = {1749--1754},
address = {Edmonton, Alberta, Canada},
month = {August}
} abstract: A Visually Guided Swimming Robot@inproceedings{chen2005auto,
title={Auto-correlation wavelet support vector machine and its applications to regression},
author={Chen, Guangyi and Dudek, Gregory},
booktitle={Proc. of Canadian Conference on Computer and Robot Vision (CRV 2005)},
pages={246--252},
year={2005},
doi={10.1109/CRV.2005.19},
address={Victoria, Canada}
} abstract: Auto-correlation wavelet support vector machine and its applications to regression@inproceedings{Rekleitis2005,
author = {Ioannis Rekleitis and Gregory Dudek},
title = {Automated Calibration of a Camera Sensor Network},
booktitle = {Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
year = {2005},
pages = {401--406},
address = {Edmonton, Alberta, Canada},
month = aug
} abstract: In this paper we present a new approach for the online calibration of a camera sensor network. This is the first step towards fully exploiting the potential for collaboration between mobile robots and static sensors sharing the same network. In particular we propose an approach for extracting the 3D pose of each camera in a common reference frame, with the help of a mobile robot. The camera poses can then be used to further refine the robot pose or to perform other tracking tasks. The analytical formulation of the problem of pose recovery@inproceedings{verma2005kinematic,
title={Kinematic Variables Estimation using Eye-in-Hand Robot Camera System},
author={Verma, Siddarth and Sharf, Inna and Dudek, Gregory},
booktitle={Proc. The 2nd Canadian Conference on Computer and Robot Vision},
pages={550--557},
year={2005},
address={Victoria, BC},
month={May}
} abstract: Vision-based motion variable estimation has been an area of intensive interest, especially for emerging applications in space robotics such as satellite maintenance, refueling and the removal of space debris. For each of these tasks, accurate kinematic motion estimates of an object are required before a robot can approach or interact with the object. In this paper, a technique is presented for autonomous identification of an object against a cluttered background and simultaneous estimation of kinematic variables of the object undergoing@inproceedings{marinakis2005learning,
title={Learning Sensor Network Topology through Monte Carlo Expectation Maximization},
author={Marinakis, Dimitri and Dudek, Gregory},
booktitle={Proc. IEEE Intl. Conf. on Robotics and Automation},
pages={4581--4587},
year={2005},
organization={IEEE},
address={Barcelona, Spain},
month={April}
} abstract: We consider the problem of inferring sensor positions and a topological (ie qualitative) map of an environment given a set of cameras with non-overlapping fields of view. In this way, without prior knowledge of the environment nor the exact position of sensors within the environment, one can infer the topology of the environment, and common traffic patterns within it. In particular, we consider sensors stationed at the junctions of the hallways of a large building. We infer the sensor connectivity graph and the travel times between sensors@article{rao2007randomized,
abstract = {The problem of minimum distance localization in environments that may contain self-similarities is addressed. A mobile robot is placed at an unknown location inside a 2 D self-similar polygonal environment P. The robot has a map of P and can compute visibility data through sensing. However, the self-similarities in the environment mean that the same visibility data may correspond to several different locations. The goal, therefore, is to determine the robot's true initial location while minimizing the distance traveled by the robot},
author = {Rao, Malvika and Dudek, Gregory and Whitesides, Sue},
journal = {The International Journal of Robotics Research},
number = {9},
pages = {917--933},
pub_year = {2007},
publisher = {Sage Publications Sage UK: London, England},
title = {Randomized algorithms for minimum distance localization},
venue = {The International Journal of …},
volume = {26}
} abstract: The problem of minimum distance localization in environments that may contain self-similarities is addressed. A mobile robot is placed at an unknown location inside a 2 D self-similar polygonal environment P. The robot has a map of P and can compute visibility data through sensing. However, the self-similarities in the environment mean that the same visibility data may correspond to several different locations. The goal, therefore, is to determine the robot's true initial location while minimizing the distance traveled by the robot@inproceedings{garden2005semantic,
title={Semantic feedback for hybrid recommendations in Recommendz},
author={Garden, Matthew and Dudek, Gregory},
booktitle={Proceedings of the IEEE International Conference on e-Technology, e-Commerce, and e-Service (EEE05)},
pages={},
year={2005},
address={Hong Kong, China},
month={March}
} abstract: In this paper we discuss the Recommendz recommender system. This domain-independent system combines the advantages of collaborative and content-based filtering in a novel way. By allowing users to provide feedback not only about an item as a whole, but also properties of an item that motivated their opinion, increased performance seems to be achieved. The features used to describe items are specified by the users of the system rather than predetermined using manual knowledge-engineering. We describe a method for combining@inproceedings{marinakis2005topology,
author = {Dimitri Marinakis and Gregory Dudek},
title = {Topology Inference for a Vision-Based Sensor Network},
booktitle = {Proc. of Canadian Conference on Computer and Robot Vision (CRV 2005)},
year = {2005},
address = {Victoria, Canada},
month = {May},
note = {Winner of best paper in the Robotics Category}
} abstract: In this paper we describe a technique to infer the topology and connectivity information of a network of cameras based on observed motion in the environment. While the technique can use labels from reliable cameras systems, the algorithm is powerful enough to function using ambiguous tracking data. The method requires no prior knowledge of the relative locations of the cameras and operates under very weak environmental assumptions. Our approach stochastically samples plausible agent trajectories based on a delay model that allows for@inproceedings{chen2005using,
title={Using Wavelets with Support Vector Machines for Recognition},
author={Chen, Guangyi and Dudek, Gregory},
booktitle={Proc. of Canadian Conference on Computer and Robot Vision (CRV 2005)},
year={2005},
address={Victoria, Canada},
month={May}
} abstract: Using Wavelets with Support Vector Machines for Recognition@inproceedings{simhon2004analogical,
author = {Saul Simhon and Gregory Dudek},
title = {Analogical Path Planning},
booktitle = {Proceedings of the National Conference on Artificial Intelligence (AAAI)},
year = {2004},
address = {San Jose, CA},
month = {July},
pages = {6}
} abstract: We present a probabilistic method for path planning that considers trajectories constrained by both the environment and an ensemble of restrictions or preferences on preferred motions for a moving robot. Our system learns constraints and preference biases on a robot's motion from examples, and then synthesizes behaviors that satisfy these constraints. This behavior can encompass motions that satisfy diverse requirements such as a sweep pattern for floor coverage, or, in particular in our experiments, satisfy restrictions on the@inproceedings{georgiades2004aqua,
title={AQUA: an aquatic walking robot},
author={Georgiades, C. and German, A. and Hogue, A. and Liu, H. and Prahacs, C. and Ripsman, A. and Sim, R. and Torres, L.-A. and Zhang, P. and Buehler, M. and Jenkin, M. and Milios, E. and Dudek, Gregory},
booktitle={Proceedings of the IEEE/RSJ/GI International Conference on Intelligent Robots and Systems (IROS)},
year={2004},
location={Sendai, Japan}
} abstract: This paper describes an underwater walking robotic system being developed under the name AQUA, the goals of the AQUA project, the overall hardware and software design, the basic hardware and sensor packages that have been developed, and some initial experiments. The robot is based on the RHex hexapod robot and uses a suite of sensing technologies, primarily based on computer vision and INS, to allow it to navigate and map clear shallow-water environments. The sensor-based navigation and mapping algorithms@inproceedings{Torres-Mendez2004,
author = {Luz-Abril Torres-Mendez and Paul Di Marco and Gregory Dudek},
title = {Inter-image Statistics for Scene Reconstruction},
booktitle = {First Canadian Conference on Computer and Robot Vision},
year = {2004},
address = {London, ON},
pages = {6}
} abstract: This paper developed prior work which incrementally completes a sparse depth map based on inter-image statistics information. In that prior work, we have observed that pixel ordering of the incremental recovery is critical to the quality of the final results. In this paper we demonstrate improved performance using an information-driven recovery policy to determine this ordering. We have also observed that the reconstruction across depth discontinuities was often problematic as there was comparatively little constraint for@inproceedings{sim2001learning,
title={Learning Generative Models of Scene Features},
author={Sim, Robert and Dudek, Gregory},
booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
year={2001},
pages={7},
address={Hawaii}
} abstract: We present a method for learning a set of generative models which are suitable for representing selected image-domain features of a scene as a function of changes in the camera viewpoint. Such models are important for robotic tasks, such as probabilistic position estimation (ie localization), as well as visualization. Our approach entails the automatic selection of the features, as well as the synthesis of models of their visual behavior. The model we propose is capable of generating maximum-likelihood views, as well as a@inproceedings{sim2004online,
title={Online Control Policy Optimization for Minimizing Map Uncertainty During Exploration},
author={Sim, Robert and Roy, Nicholas and Dudek, Gregory},
booktitle={Proceedings of the IEEE International Conference on Robotics and Automation (ICRA)},
pages={6},
year={2004},
address={New Orleans, LA},
month={April}
} abstract: Tremendous progress has been made recently in simultaneous localization and mapping of unknown environments. Using sensor and odometry data from an exploring mobile robot, it has become much easier to build high-quality globally consistent maps of many large, real-world environments. To date, however, relatively little attention has been paid to the controllers used to build these maps. Existing exploration strategies usually attempt to cover the largest amount of unknown space as quickly as possible. Few strategies exist for@inproceedings{simhon2004pen,
author = {Saul Simhon and Gregory Dudek},
title = {Pen Stroke Extraction and Refinement using Learned Models},
booktitle = {Proc. Eurographics Workshop on Sketch-Based Interfaces and Modeling},
year = {2004},
address = {Grenoble, France},
month = {August}
} abstract: This paper presents a smart interface that automatically extracts and refines pen strokes from images of hand drawn sketches. The interface allows users to digitize hand-drawn material such sketches of flowcharts, cartoons or other pen based drawings and automatically isolate and refine the individual strokes making up the sketch. First, we present a method for extracting pen strokes based on learned constraints on curves. The approach consists of using a training set that shows good examples of curves and how a user would@inproceedings{bourque2004procedural,
author = {Eric Bourque and Gregory Dudek},
title = {Procedural Texture Matching and Transformation},
booktitle = {Proc. Eurographics 2004},
year = {2004},
month = {August},
pages = {8},
journal = {Computer Graphics Forum}
} abstract: We present a technique for creating a smoothly varying sequence of procedural textures that interpolates between arbitrary input samples of texture. This texture transformation uses a library of procedural shaders and selects the correct shaders and associated parameters to accomplish the task. In general, selecting a procedural texture from a library, or finding the correct parameters to produce a smooth texture transition can be complex and time consuming. We propose a strategy for automating this process. While superficially this@article{rao2007randomized,
abstract = {The problem of minimum distance localization in environments that may contain self-similarities is addressed. A mobile robot is placed at an unknown location inside a 2 D self-similar polygonal environment P. The robot has a map of P and can compute visibility data through sensing. However, the self-similarities in the environment mean that the same visibility data may correspond to several different locations. The goal, therefore, is to determine the robot's true initial location while minimizing the distance traveled by the robot},
author = {Rao, Malvika and Dudek, Gregory and Whitesides, Sue},
journal = {The International Journal of Robotics Research},
number = {9},
pages = {917--933},
pub_year = {2007},
publisher = {Sage Publications Sage UK: London, England},
title = {Randomized algorithms for minimum distance localization},
venue = {The International Journal of …},
volume = {26}
} abstract: The problem of minimum distance localization in environments that may contain self-similarities is addressed. A mobile robot is placed at an unknown location inside a 2 D self-similar polygonal environment P. The robot has a map of P and can compute visibility data through sensing. However, the self-similarities in the environment mean that the same visibility data may correspond to several different locations. The goal, therefore, is to determine the robot's true initial location while minimizing the distance traveled by the robot@inproceedings{Torres-Mendez2004,
author = {Luz-Abril Torres-Mendez and Gregory Dudek},
title = {Reconstruction of 3D Models from Intensity Images and Partial Depth},
booktitle = {Proceedings of the National Conference on Artificial Intelligence (AAAI)},
year = {2004},
address = {San Jose, CA},
month = {July},
pages = {6}
} abstract: This paper addresses the probabilistic inference of geometric structures from images. Specifically, of synthesizing range data to enhance the reconstruction of a 3D model of an indoor environment by using video images and (very) partial depth information. In our method, we interpolate the available range data using statistical inferences learned from the concurrently available video images and from those (sparse) regions where both range and intensity information is available. The spatial relationships between the variations in intensity@inproceedings{sim2004self,
title={Self-Organizing Visual Maps},
author={Sim, Robert and Dudek, Gregory},
booktitle={Proceedings of the National Conference on Artificial Intelligence (AAAI)},
pages={6},
year={2004},
address={San Jose, CA},
month={July}
} abstract: This paper deals with automatically learning the spatial distribution of a set of images. That is, given a sequence of images acquired from well-separated locations, how can they be arranged to best explain their genesis? The solution to this problem can be viewed as an instance of robot mapping although it can also be used in other contexts. We examine the problem where only limited prior odometric information is available, employing a feature-based method derived from a probabilistic pose estimation framework. Initially, a set of@inproceedings{Simhon2004,
author = {Saul Simhon and Gregory Dudek},
title = {Sketch Interpretation and Refinement using Statistical Models},
booktitle = {Proc. Eurographics Symposium on Rendering},
year = {2004}
} abstract: We present a system for generating 2D illustrations from hand drawn outlines consisting of only curve strokes. A user can draw a coarse sketch and the system would automatically augment the shape, thickness, color and surrounding texture of the curves making up the sketch. The styles for these refinements are learned from examples whose semantics have been pre-classified. There can be several styles applicable on a curve and the system automatically identifies which one to use and how to use it based on a curve's shape and its@inproceedings{Torres-Mendez2004,
author = {Luz Abril Torres-Mendez and Gregory Dudek},
title = {Statistical Inference and Synthesis in the Image Domain for Mobile Robot Environment Modeling},
booktitle = {Proceedings of the IEEE/RSJ/GI International Conference on Intelligent Robots and Systems (IROS)},
year = {2004},
address = {Sendai, Japan}
} abstract: We address the problem of computing dense range maps of indoor locations using only intensity images and partial depth. We allow a mobile robot to navigate the environment, take some pictures and few range data. Our method is based on interpolating the existing range data using statistical inferences learned from the available intensity image and from those (sparse) regions where both range and intensity information is present. The spatial relationships between the variations in intensity and range can be efficiently captured by the@inproceedings{Torres-Mendez2004,
author = {Luz Abril Torres-Mendez and Gregory Dudek},
title = {Statistics in the Image Domain for Mobile Robot Environment Modeling},
booktitle = {Proc. 4th International Symposium of Robotics and Automation},
year = {2004},
address = {Queretaro, Mexico},
month = {August}
} abstract: This paper addresses the problem of estimating dense range maps of indoor locations using only intensity images and sparse partial depth information. Unlike shape-fromshading, we infer the relationship between intensity and range data and use it to produce a complete depth map. We extend prior work by incorporating geometric information from the available range data, specifically, we add surface normal information to reconstruct surfaces whose variations are not captured in the initial range measurements. In addition, the order on which@inproceedings{sim2003comparing,
title={Comparing image-based localization methods},
author={Sim, Robert and Dudek, Gregory},
booktitle={Proc. of the International Joint Conference on Artificial Intelligence (IJCAI)},
pages={3},
year={2003},
address={Acapulco, Mexico},
month={Aug}
} abstract: This paper compares alternative approaches to pose estimation using visual cues from the environment. We examine approaches that derive pose estimates from global image properties, such as principal components analysis (PCA) versus from local image properties, commonly referred to as landmarks. We also consider the failure-modes of the different methods. Our work is validated with experimental results.@inproceedings{sim2003effective,
title={Effective Exploration Strategies for the Construction of Visual Maps},
author={Sim, Robert and Dudek, Gregory},
booktitle={Proceedings of the IEEE/RSJ Conference on Intelligent Robots and Systems (IROS)},
pages={8},
year={2003},
address={Las Vegas, NV},
month={Oct}
} abstract: We consider the effect of exploration policy in the context of the autonomous construction of a visual map of an unknown environment. Like other concurrent mapping and localization (CML) tasks, odometric uncertainty poses the problem of introducing distortions into the map which are difficult to correct without costly on-line or post-processing algorithms. Our problem is further compounded by the implicit nature of the visual map representation, which is designed to accommodate a wide variety of visual phenomena without assuming a@inproceedings{sim2003examining,
title={Examining Exploratory Trajectories for Minimizing Map Uncertainty},
author={Sim, Robert and Dudek, Gregory},
booktitle={Proceedings of the International Joint Conference on Artificial Intelligence (IJCAI) workshop on Reasoning with Uncertainty in Robotics (RUR)},
pages={8},
year={2003},
address={Acapulco, Mexico},
month={Aug}
} abstract: We examine the problem of minimizing uncertainty in the automated construction of a visual map of an unknown environment. Our work is motivated by the idea that a robot's exploration policy can impact the accuracy of the resulting map, and we seek to examine the behavior of a set of policies that exhibit a trade-off between accuracy and efficiency. We are further motivated by the specific requirements of our map representation, which learns a set of implicit models of visual features. Such a representation precludes the instantiation of@inproceedings{Rekleitis2003,
author = {Ioannis Rekleitis and Evangelos Milios and Gregory Dudek},
title = {Experiments in Free-Space Triangulation Using Cooperative Localization},
booktitle = {Proceedings of the IEEE/RSJ Conference on Intelligent Robots and Systems (IROS)},
year = {2003},
address = {Las Vegas, NV},
month = {Oct.},
pages = {8}
} abstract: This paper presents a first detailed case study of collaborative exploration of a substantial environment. We use a pair of cooperating robots to test multi-robot environment mapping algorithms based on triangulation of free space. The robots observe one another using a robot tracking sensor based on laser range sensing (LIDAR). The environment mapping itself is accomplished using sonar sensing. The results of this mapping are compared to those obtained using scanning laser range sensing and the scan matching algorithm. We@inproceedings{Garden2003,
author = {Matthew Garden and Gregory Dudek},
title = {On User Recommendations Based on Multiple Cues},
booktitle = {Proc. IEEE WI/IAT 2003 Workshop on Applications, Products and Services of Web-Based Support Systems (in conjunction with IEEE/WIC International Conference on Web Intelligence)},
year = {2003},
pages = {139--144},
address = {Halifax, NS},
isbn = {0-9734039-1-8}
} abstract: In this paper we present an overview of a recommender system that attempts to predict user preferences based on several sources including prior choices and selected user-defined features. By using a combination of collaborative filtering and semantic features, we hope to provide performance superior to either alone. Further, our set of semantic features is acquired and updated using a learning-based procedure that avoids the need for manual knowledgeengineering. Our system is implemented in a web-based application server@inproceedings{simhon2003path,
author = {Saul Simhon and Gregory Dudek},
title = {Path Planning Using Learned Constraints and Preferences},
booktitle = {IEEE International Conference on Robotics and Automation},
year = {2003},
address = {Taipei, Taiwan},
month = {May}
} abstract: In this paper we present a novel method for robot path planning based on learning motion patterns. A motion pattern is defined as the path that results from applying a set of probabilistic constraints to a" raw" input path. For example, a user can sketch an approximate path for a robot without considered issues such as bounded radius of curvature and our system would then elaborate it to include such a constraint. In our approach, the constraints that generate a path are learned by capturing the statistical properties of a set of@inproceedings{Rekleitis2003,
author = {Ioannis Rekleitis and Robert Sim and Evangelos Milios and Gregory Dudek},
title = {Probabilistic Cooperative Localization and Mapping in Practice},
booktitle = {IEEE International Conference on Robotics and Automation},
year = {2003},
address = {Taipei, Taiwan},
month = {May}
} abstract: In this paper we present a probabilistic framework for the reduction in the uncertainty of a moving robot pose during exploration by using a second robot to assist. A Monte Carlo Simulation technique (specifically, a Particle Filter) is employed in order to model and reduce the accumulated odometric error. Furthermore, we study the requirements to obtain an accurate yet timely pose estimate. A team of two robots is employed to explore an indoor environment in this paper, although several aspects of the approach have been extended to@inproceedings{Torres-Mendez2003,
author = {Luz-Abril Torres-Mendez and Gregory Dudek},
title = {Range Synthesis for 3D Environment Modeling},
booktitle = {Proceedings of the IEEE/RSJ Conference on Intelligent Robots and Systems (IROS)},
address = {Las Vegas, NV},
month = {October},
year = {2003},
pages = {8}
} abstract: In this paper a range synthesis algorithm is proposed as an initial solution to the problem of 3D environment modeling from sparse data. We develop a statistical learning method for inferring and extrapolating range data from as little as one intensity image and from those (sparse) regions where both range and intensity information is available. Our work is related to methods for texture synthesis using Markov Random Field methods. We demonstrate that MRF methods can also be applied to general intensity images with little associated range@inproceedings{sim2003robodaemon,
title={RoboDaemon - A device-independent, network-oriented, modular mobile robot controller},
author={Sim, Robert and Dudek, Gregory},
booktitle={IEEE International Conference on Robotics and Automation},
year={2003},
address={Taipei, Taiwan},
month={May}
} abstract: We discuss a software environment for multi-robot, multi-platform mobile robot control and simulation. Like others, we have observed that mobile robotics research is greatly facilitated by the availability of a suitable simulator for both vehicle kinematics as well as sensing, and have created an environment that permits this while allowing a large measure of device independence. By using a multiprocessor internet-based architecture, our platform permits multiple users to use a variety of programming interfaces (visual, script-based or various@inproceedings{Torres-Mendez2002,
author = {L. A. Torres-Mendez and Gregory Dudek},
title = {Automated Enhancement of 3-D Models},
booktitle = {Proc. Eurographics 2002 (Geometric and Physics Based Modeling)},
address = {Saarbrucken, Germany},
month = {September},
year = {2002}
} abstract: The acquisition of a 3D model of a real environment can be accomplished using range sensors. In practice, suitable sensors to densely cover a large environment are often impractical. This paper presents ongoing work on the synthesis of 3D environment models from as little as one intensity image and sparse range data. Our method is based on interpolating the available range data using statistical inferences learned from the available intensity image and from those (sparse) regions where both range and intensity information@inproceedings{bourque2002automated,
title={Automated Parameter Estimation for Procedural Texturing},
author={Bourque, Eric and Dudek, Gregory},
booktitle={Proc. 13th Eurographics Workshop on Rendering},
year={2002},
location={Pisa, Italy}
} abstract: @inproceedings{Rekleitis2002,
author = {Ioannis Rekleitis and Robert Sim and Evangelos Milios and Gregory Dudek},
title = {Multi-Robot Cooperative Localization: A Study of Trade-offs Between Efficiency and Accuracy},
booktitle = {Proc. IEEE/RSJ International Conference on Intelligent Robots and Systems},
year = {2002},
pages = {2690--2695},
address = {Lausanne, Switzerland}
} abstract: This paper examines the tradeoffs between different classes of sensing strategy and motion control strategy in the context of terrain mapping with multiple robots. We consider a larger group of robots that can mutually estimate one another's position (in 2D or 3D) and uncertainty using a sample-based (particle filter) model of uncertainty. Our prior work has dealt with a pair of robots that estimate one another's position using visual tracking and coordinated motion. Here we extend these results and consider a richer set of sensing and@inproceedings{Rekleitis2002,
author = {Ioannis Rekleitis and Robert Sim and Evangelos Milios and Gregory Dudek},
title = {On the Positional Uncertainty of Multi-Robot Cooperative Localization},
booktitle = {Proc. Naval Research Labs/NATO Workshop on Multi-Robot Systems},
year = {2002},
address = {Washington, DC},
note = {to appear}
} abstract: This paper deals with terrain mapping and position estimation using multiple robots. Here we will discuss work where a larger group of robots can mutually estimate one another's position (in 2D or 3D) and uncertainty using a sample-based (particle filter) model of uncertainty. Our prior work has dealt with a pair of robots that estimate one another's position using visual tracking and coordinated motion and we extend these results and consider a richer set of sensing and motion options. In particular, we focus on issues related to@inproceedings{Torres-Mendez2003,
author = {Luz-Abril Torres-Mendez and Gregory Dudek},
title = {Range Synthesis for 3D Environment Modeling},
booktitle = {Proceedings of the IEEE/RSJ Conference on Intelligent Robots and Systems (IROS)},
address = {Las Vegas, NV},
month = {October},
year = {2003},
pages = {8}
} abstract: In this paper a range synthesis algorithm is proposed as an initial solution to the problem of 3D environment modeling from sparse data. We develop a statistical learning method for inferring and extrapolating range data from as little as one intensity image and from those (sparse) regions where both range and intensity information is available. Our work is related to methods for texture synthesis using Markov Random Field methods. We demonstrate that MRF methods can also be applied to general intensity images with little associated range@inproceedings{Rekleitis2001,
author = {Ioannis Rekleitis and Robert Sim and Evangelos Milios and Gregory Dudek},
title = {Collaborative Exploration for Map Construction},
booktitle = {Proc. International Symposium on Computational Intelligence in Robotics and Automation (CIRA)},
year = {2001},
location = {Banff, Canada},
pages = {6}
} abstract: We consider the problem of map learning while maintaining ground-truth pose estimates. Map learning is important in tasks that require a model of the environment or some of its features. As a robot collects data, uncertainty about its position accumulates and corrupts its knowledge of the positions from which observations are taken. We address this problem by employing cooperative localization; that is, deploying a second robot to observe the other as it explores, thereby establishing a virtual tether, and enabling an accurate estimate of the@inproceedings{Rekleitis2001,
author = {Ioannis Rekleitis and Robert Sim and Evangelos Milios and Gregory Dudek},
title = {Collaborative Exploration for the Construction of Visual Maps},
booktitle = {Proceedings of IEEE/RSJ Conference on Intelligent Robots and Systems (IROS)},
year = {2001},
location = {Hawaii},
pages = {6}
} abstract: We examine the problem of learning a visual map of the environment while maintaining an accurate pose estimate. Our approach is based on using two robots in a simple collaborative scheme. Without outside information, as a robot collects training images, its position estimate accumulates errors, thus corrupting its knowledge of the positions from which observations are taken. We address this problem by deploying a second robot to observe the first one as it explores, thereby establishing a virtual tether, and enabling an accurate@article{roy2001collaborative,
title={Collaborative Robot Exploration and Rendezvous: Algorithms, Performance Bounds and Observations},
author={Roy, Nicholas and Dudek, Gregory},
journal={Autonomous Robots},
volume={11},
number={2},
pages={117--136},
year={2001}
} abstract: We consider the problem of how two heterogeneous robots can arrange to meet in an unknown environment from unknown starting locations: that is, the problem of arranging a robot rendezvous. We are interested, in particular, in allowing two robots to rendezvous so that they can collaboratively explore an unknown environment. Specifically, we address the problem of how a pair of exploring agents that cannot communicate with one another over long distances can meet if they start exploring at different unknown locations in an unknown@inproceedings{bourque2001image,
author = {Eric Bourque and Gregory Dudek},
title = {Image-Driven Procedural Texture Specification},
booktitle = {Proc. Vision Interface},
year = {2001},
pages = {1--6},
address = {Ottawa, Canada},
month = {June}
} abstract: In this paper we describe an approach to the automated specification of procedural textures to be used in rendering, based on representative samples. Procedural textures exhibit many advantages over traditional surface texturing techniques, but unfortunately finding the correct procedural texture and appropriate parameters to create the desired texture can be a daunting task for even the most experienced computer graphics artists. The method we propose here, which we refer to as image-based procedural texturing, allows the@inproceedings{sim1999learning,
author = {Robert Sim and Gregory Dudek},
title = {Learning Environmental Features for Position Estimation},
booktitle = {Proc. of the IEEE Workshop on Perception for Mobile Agents},
year = {1999},
pages = {7--14},
address = {Fort Collins, Colorado},
month = {June}
} abstract: We present a method for learning a set of environmental features which are useful for pose estimation. The landmark learning mechanism is designed to be applicable to a wide range of environments, and generalized for different sensing modalities. In the context of computer vision, each landmark is detected as a local extremum of a measure of distinctiveness and represented by an appearance-based encoding which is exploited for matching. The set of obtained landmarks can be parameterized and then evaluated in terms of their utility for the@inproceedings{sim2001learning,
title={Learning Generative Models of Scene Features},
author={Sim, Robert and Dudek, Gregory},
booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
year={2001},
pages={7},
address={Hawaii}
} abstract: We present a method for learning a set of generative models which are suitable for representing selected image-domain features of a scene as a function of changes in the camera viewpoint. Such models are important for robotic tasks, such as probabilistic position estimation (ie localization), as well as visualization. Our approach entails the automatic selection of the features, as well as the synthesis of models of their visual behavior. The model we propose is capable of generating maximum-likelihood views, as well as a@inproceedings{Rekleitis2000,
author = {Ioannis Rekleitis and Gregory Dudek},
title = {Multi-Robot Collaboration for Robust Exploration},
booktitle = {Proc. of IEEE International Conference in Robotics and Automation},
year = {2000},
pages = {3164--3169},
address = {San Francisco, CA},
month = apr,
} abstract: This paper presents a new sensing modality for multirobot exploration. The approach is based on using a pair of robots that observe each other, and act in concert to reduce odometry errors. We assume the robots can both directly sense nearby obstacles and see each other. The proposed approach improves the quality of the map by reducing the inaccuracies that occur over time from dead reckoning errors. Furthermore, by exploiting the ability of the robots to see each other, we can detect opaque obstacles in the environment@inproceedings{Simhon2001,
author = {Simhon, S. and Gregory Dudek},
title = {On the Elaboration of Hand-Drawn Sketches},
booktitle = {Active Media Technology},
year = {2001},
address = {Hong Kong},
month = {Dec}
} abstract: This work considers an approach for artificially enhancing the richness and level of detail of graphical scenes. In particular, we examine a method for automatically generating high-resolution novel curves from manually sketched drawings of those curves. The essential idea is to augment the hand-drawn curves using prior knowledge to produce a more elaborated picture. Our method uses multi-scale analysis of a class of training data to capture statistical properties of the set. These properties are then conditioned at a coarse@article{Polifroni2000,
author = {Sandra Polifroni and Frank Ferrie and Gregory Dudek},
title = {Evaluation of Computation Attention Operators using Human Image Recognition},
journal = {Investigative Opthalmology and Visual Science (Suppl)},
year = {2000},
month = {May},
pages = {196--197},
publisher = {The Association for Research in Vision and Opthalmology}
} abstract: This thesis presents a novel method of evaluating computationaI attention operators, which select locations of interest in an image, using a human image recognition task. Assuming that locations which are maximally interesting will be most useful for recognizing an image, it follows that a location selected by an attention operator will facilitate image recognition if it is of interest to a human. Since attention operators are increasingly being used to replace humans in vision tasks, it is relevant that their performance be compared to human vision@inproceedings{Rekleitis2000,
author = {Ioannis Rekleitis and Evangelos Milios and Gregory Dudek},
title = {Graph-Based Exploration using Multiple Robots},
booktitle = {Proc. 5th International Symposium on Distributed Autonomous Robotic Systems (DARS)},
year = {2000},
pages = {241--250},
address = {Knoxville, USA},
month = {October}
} abstract: We present an approach to multi-robot exploration of large environments. Our method is designed to be robust in the face of arbitrarily large odometry errors or objects with poor reflectance characteristics. The algorithm achieves its robustness by using a team of cooperating agents. The critical aspect of our method is the use of a vision system that sweeps areas of free space and generates a graph-based description of the environment. This graph is used to guide the exploration process and can also be used for subsequent@inproceedings{jugessur2000local,
title={Local Appearance for Robust Object Recognition},
author={Jugessur, Deeptiman and Dudek, Gregory},
booktitle={Proc. IEEE Computer Vision and Pattern Recognition},
year={2000},
month={June}
} abstract: We present an approach to appearance-based object recognition using single camera images. Our approach is based on using an attention mechanism to obtain visual features that are generic, robust and informative. The features themselves are recognized using principal components an the frequency domain. In this paper we show how the visual characteristics of only a small number of such features can be used for appearance-based object recognition that is not confounded by planar rotations or background clutter.@inproceedings{Rekleitis2000,
author = {Ioannis Rekleitis and Gregory Dudek},
title = {Multi-Robot Collaboration for Robust Exploration},
booktitle = {Proc. of IEEE International Conference in Robotics and Automation},
year = {2000},
pages = {3164--3169},
address = {San Francisco, CA},
month = apr,
} abstract: This paper presents a new sensing modality for multirobot exploration. The approach is based on using a pair of robots that observe each other, and act in concert to reduce odometry errors. We assume the robots can both directly sense nearby obstacles and see each other. The proposed approach improves the quality of the map by reducing the inaccuracies that occur over time from dead reckoning errors. Furthermore, by exploiting the ability of the robots to see each other, we can detect opaque obstacles in the environment@article{bourque2000automated,
title={On the Automated Construction of Image-Based Maps},
author={Bourque, Eric and Dudek, Gregory},
journal={Autonomous Robots},
volume={8},
number={2},
pages={103--104},
year={2000},
month={April}
} abstract: For many tasks, we wish to record or recover the description of a remote environment so that it can be inspected by a person. This is the problem we address in this paper. Rather than recovering a geometric description of an environment, as many robotics systems attempt to do, we seek to recover a model of an environment in terms of its appearance from a set of carefully selected viewpoints. Our hope is that this type of model is both more accessible to humans for many realistic tasks, and also more readily achieved with automated systems@inproceedings{bourque2000online,
author = {Eric Bourque and Gregory Dudek},
title = {On-line Construction of Iconic Maps},
booktitle = {Proc. of IEEE International Conference in Robotics and Automation},
year = {2000},
address = {San Francisco, CA},
month = {April}
} abstract: This paper describes an approach to the automated creation of virtual realities (or virtual maps) of an a priori unknown environment by using a mobile robot. The method we propose is aimed at the creation of an image-based or iconic map, rather than a representation in terms of 2D or 3D spatial occupancy. A key aspect of this is having a mobile robot automatically select points and views of interest that can be used to exemplify the appearance of the environment. This paper develops the use of alpha-backtracking as a@inproceedings{jugessur2000robust,
title={Robust Place Recognition using Local Appearance-Based Methods},
author={Jugessur, Deeptiman and Dudek, Gregory},
booktitle={Proc. of IEEE International Conference in Robotics and Automation},
year={2000},
month={April},
address={San Francisco, CA}
} abstract: We present an approach to the automatic recognition of locations or landmarks using single camera images. Our approach is to learn visual features in the appearance domain that can be used to characterize an object or a location. These features are defined statistically and then are recognized using principal components in the frequency domain. We show that this technique can be used to recognize specific objects on varying backgrounds, as well as environmental features.@inproceedings{simhon2000stochastic,
author = {Saul Simhon and Gregory Dudek},
title = {Stochastic Reconstruction from Coarse Data},
booktitle = {Proc. Graphics Interface},
year = {2000},
address = {Montreal},
month = {May},
pages = {120--126}
} abstract: Stochastic Reconstruction from Coarse Data@inproceedings{jenkin2000paparazzi,
author = {Michael Jenkin and Gregory Dudek},
title = {The Paparazzi Problem},
booktitle = {Proc. IEEE/RSJ IROS 2000},
year = {2000},
address = {Takamatsu, Japan}
} abstract: Multiple mobile robots, or robot collectives, have been proposed as solutions to various tasks in which distributed sensing and action are required. Here we consider applying a collective of robots to the paparazzi problem-the problem of providing sensor coverage of a target robot. We demonstrate how the computational task of the collective can be formulated as a global energy minimization task over the entire collective, and show how individual members of the collective can solve the task in a distributed fashion so that the entire@book{dudek2000computational,
title={Computational Principles of Mobile Robotics (first edition)},
author={Dudek, Gregory and Jenkin, Michael},
edition={1st},
year={2000},
publisher={Cambridge University Press},
isbn={9780521692120},
pages={450}
} abstract: Computational Principles of Mobile Robotics (first edition)@inproceedings{Rekleitis1999,
author = {Ioannis Rekleitis and Vida Dujmovi\'c and Gregory Dudek},
title = {Efficient Topological Exploration},
booktitle = {Proc. of IEEE International Conference in Robotics and Automation},
year = {1999},
pages = {676--681},
address = {Detroit, MI},
month = {May}
} abstract: We consider the robot exploration of a planar graph-like world. The robot's goal is to build a complete map of its environment. The environment is modeled as an arbitrary undirected planar graph which is initially unknown to the robot. The robot cannot distinguish vertices and edges that it has explored from the unexplored ones. The robot is assumed to be able to autonomously traverse graph edges, recognize when it has reached a vertex, and enumerate edges incident upon the current vertex. The robot cannot measure distances nor@article{Badra1999,
author = {F. Badra and Q. Qumsieh and Gregory Dudek},
title = {Image Mosaicking Using Zernike Moments},
journal = {International Journal of Pattern Recognition and Artificial Intelligence (IJPRAI)},
volume = {13},
number = {4},
pages = {685--704},
month = {August},
year = {1999}
} abstract: Image Mosaicking Using Zernike Moments@inproceedings{sim1999learning,
author = {Robert Sim and Gregory Dudek},
title = {Learning and Evaluating Visual Features for Pose Estimation},
booktitle = {Proceedings of the International Conference on Computer Vision},
year = {1999},
pages = {32--40},
address = {Kerkyra (Corfu), Greece},
month = sep
} abstract: We present a method for learning a set of visual landmarks which are useful for pose estimation. The landmark learning mechanism is designed to be applicable to a wide range of environments, and generalized for different approaches to computing a pose estimate. Initially, each landmark is detected as a focal extremum of a measure of distinctiveness and represented by a principal components encoding which is exploited for matching. Attributes of the observed landmarks can be parameterized using a generic parameterization method@inproceedings{sim1999learning,
author = {Robert Sim and Gregory Dudek},
title = {Learning Environmental Features for Position Estimation},
booktitle = {Proc. of the IEEE Workshop on Perception for Mobile Agents},
year = {1999},
pages = {7--14},
address = {Fort Collins, Colorado},
month = {June}
} abstract: We present a method for learning a set of environmental features which are useful for pose estimation. The landmark learning mechanism is designed to be applicable to a wide range of environments, and generalized for different sensing modalities. In the context of computer vision, each landmark is detected as a local extremum of a measure of distinctiveness and represented by an appearance-based encoding which is exploited for matching. The set of obtained landmarks can be parameterized and then evaluated in terms of their utility for the@article{sim1999learning,
title={Learning Visual Landmarks for Pose Estimation},
author={Sim, Robert and Dudek, Gregory},
journal={Canadian Artificial Intelligence},
volume={43},
year={1999},
pages={13--17}
} abstract: We present an approach to vision-based mobile robot localization, even without an a-priori pose estimate. This is accomplished by learning a set of visual features called image-domain landmarks. The landmark learning mechanism is designed to be applicable to a wide range of environments. Each landmark is detected as a focal extremum of a measure of uniqueness and represented by an appearance-based encoding. Localization is performed using a method that matches observed landmarks to learned prototypes and generates@article{sim1999learning,
title={Learning Visual Landmarks for Pose Estimation (journal version)},
author={Sim, Robert and Dudek, Gregory},
journal={Canadian Artificial Intelligence},
volume={43},
year={1999},
pages={13--17}
} abstract: Learning Visual Landmarks for Pose Estimation (journal version)@inproceedings{simhon1998global,
author = {Saul Simhon and Gregory Dudek},
title = {A global topological map formed by local metric maps},
booktitle = {Proceedings IEEE/RSJ Int. Conf. on Intelligent Robots and Systems (IROS)},
year = {1998},
pages = {1708--1714},
address = {Victoria, BC},
month = {October}
} abstract: We describe a method of mapping large scale static environments using a hybrid topological-metric model. A global map is formed from a set of local maps organized in a topological structure. Each local map contains quantitative environment information using a local reference frame. They are denoted as islands of reliability because they provide accurate metric information of the environment. The mapping problem then becomes where to place the islands of reliability and to what extent should they cover the environment. This is@article{lecours1998semantic,
abstract = {},
author = {Lecours, S and Arguin, M and Bub, D and Dudek, G and Caille, S and Fontaine, S},
journal = {BRAIN AND COGNITION},
number = {1},
pages = {138--141},
pub_year = {1998},
publisher = {ACADEMIC PRESS INC 525 B ST, STE 1900, SAN DIEGO, CA 92101-4495 USA},
title = {A semantic proximity effect on object recognition in visual agnosia for biological kinds},
venue = {… COGNITION},
volume = {37}
} abstract: @inproceedings{bourque1998automated,
author = {Eric Bourque and Gregory Dudek},
title = {Automated Image-Based Mapping},
booktitle = {Proc. IEEE Workshop on Perception for Mobile Agents},
year = {1998},
pages = {61--70},
address = {Santa Barbara, CA},
month = jun
} abstract: We describe an approach to the automated construction of visual maps of an unknown environment. These maps take the form of image-based “walk-throughs” rather than 2D or 3D models. Our approach is based on the selection of informative viewpoints within the environment. These viewpoints are locations in the environment associated with views containing maximal visual interest. This approach to environment representation is analogous to image compression. Our goal is to obtain a set of representative views@article{dudek1998localizing,
abstract = {We consider the problem of localizing a robot in a known environment modeled by a simple polygon P. We assume that the robot has a map of P but is placed at an unknown location inside P. From its initial location, the robot sees a set of points called the visibility polygon V of its location. In general, sensing at a single point will not suffice to uniquely localize the robot, since the set H of points in P with visibility polygon V may have more than one element. Hence, the robot must move around and use range sensing and a compass to},
author = {Dudek, Gregory and Romanik, Kathleen and Whitesides, Sue},
journal = {SIAM Journal on Computing},
number = {2},
pages = {583--604},
pub_year = {1998},
publisher = {SIAM},
title = {Localizing a robot with minimum travel},
venue = {SIAM Journal on Computing},
volume = {27}
} abstract: We consider the problem of localizing a robot in a known environment modeled by a simple polygon P. We assume that the robot has a map of P but is placed at an unknown location inside P. From its initial location, the robot sees a set of points called the visibility polygon V of its location. In general, sensing at a single point will not suffice to uniquely localize the robot, since the set H of points in P with visibility polygon V may have more than one element. Hence, the robot must move around and use range sensing and a compass to@inproceedings{sim1998mobile,
author = {Robert Sim and Gregory Dudek},
title = {Mobile robot localization from learned landmarks},
booktitle = {Proceedings IEEE/RSJ Int. Conf. on Intelligent Robots and Systems (IROS)},
year = {1998},
pages = {1060--1065},
address = {Victoria, BC},
month = {October}
} abstract: Presents an approach to vision-based mobile robot localization. In an attempt to capitalize on the benefits of both image and landmark-based methods, we describe a method that combines their strengths. Images are encoded as a set of visual features called landmarks. Potential landmarks are detected using an attention mechanism implemented as a measure of uniqueness. They are then selected and represented by an appearance-based encoding. Localization is performed using a landmark tracking and interpolation method which obtains@inproceedings{Daum1998,
author = {M. Daum and G. Dudek},
title = {On 3-D surface reconstruction using shape from shadows},
booktitle = {Proceedings. 1998 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (Cat. No.98CB36231)},
year = {1998},
pages = {461--468},
doi = {10/b8wprs},
month = jun
} abstract: In this paper we discuss new results on the Shape From Darkness problem: using the motion of cast shadows to recover scene structure. Our approach is based on collecting a set of images from a fixed viewpoint as a known light source mover;" across the sky". Previously published solutions to this problem have performed the reconstruction only for cross sections of the scene. In this paper, we present a reconstruction algorithm and discuss the reconstruction of an entire 3-D scene under various light source trajectories. We also@inproceedings{Rekleitis1998,
author = {Ioannis Rekleitis and Evangelos Milios and Gregory Dudek},
title = {On Multiagent Exploration},
booktitle = {Proc. Vision Interface},
year = {1998},
pages = {455--461},
address = {Vancouver, BC},
month = {June}
} abstract: This paper describes a technique for multi-agent exploration of an unknown environment, that improves the quality of the map by reducing the inaccuracies that occur over time from dead reckoning errors. We present an algorithmic solution, simulation results, as well as a cost analysis and experimental data. The approach is based on using a pair of robots that observe one another's behaviour, thus greatly reducing odometry errors. We assume the robots can both directly sense nearby obstacles and see one another. We have@inproceedings{dudek1998integration,
author = {Gregory Dudek},
title = {On the Integration of Mobile Robot Systems},
booktitle = {AAAI Spring Symposium},
year = {1998},
pages = {20--27},
address = {Stanford, CA}
} abstract: On the Integration of Mobile Robot Systems@inproceedings{daum1998out,
title={Out of the Dark: Using Shadows to Reconstruct 3D Surfaces},
author={Daum, Michael and Dudek, Gregory},
booktitle={Proc. Asian Conference on Computer Vision},
pages={72--79},
year={1998},
address={Hong Kong, China},
month={January},
series={Lecture Notes in Computer Science},
volume={1351},
editor={Goos, G. and Hartmanis, J. and van Leeuwen, J.},
publisher={Springer}
} abstract: Shape From Darkness refers to using the shadows cast by a scene to reconstruct the structure of the scene. A collection of images associated with different light source positions is used. Previously published solutions to this problem have performed the reconstruction only for cross sections of the scene. We propose a variant of Shape From Darkness which is capable of reconstructing the entire 3-D scene. In addition, this algorithm can be applied to a broader class of light source trajectories, including trajectories which@article{dudek1998profile,
author = {Dudek, G.},
title = {Profile of a Winner: McGill University},
journal = {American Association for Artificial Intelligence Magazine},
year = {1998},
month = {Summer},
note = {(part of an article of the AAAI-97 mobile robotics competition)}
} abstract: Much of the work at the McGill Mobile Robotics Lab concerns computational problems related to use of sensors: vision, laser, and sonar. As a result, the McGill team entered the nonmanipulator category. The team was made up of four students: Francois Belair, Eric Bourque, Deeptiman Jugessur, and Robert Sim, with myself acting as faculty mentor. The robot they used was a NOMAD 200, a chest-height cylindrical robot with a three-wheeled synchrodrive, a standard ring of 16 sonar sensors, and a single-color camera mounted on a@inproceedings{bourque1998robotic,
author = {Eric Bourque and Philippe Ciaravola and Gregory Dudek},
title = {Robotic Sightseeing - A Method for Automatically Creating Virtual Environments},
booktitle = {Proc. IEEE International Conference on Robotics and Automation},
year = {1998},
pages = {3186--3191},
publisher = {IEEE Press},
address = {Leuven, Belgium},
month = {May}
} abstract: This paper describes the fully automatic creation of an environment's description using an image-based representation. This representation is a collection of cylindrical sample images combined into an" image-based virtual reality". The locations at which the environment will be sampled are chosen automatically using an operator inspired by models of human visual attention and saccadic motion. The image acquisition is performed by a mobile robot. The selection of vantage points is based on an analysis of the edge structure of sampled@inproceedings{dudek1998robotics,
author = {Gregory Dudek},
title = {Robotics and Empiricism},
booktitle = {AAAI Spring Symposium (position paper)},
year = {1998},
pages = {97--98},
address = {Stanford, CA}
} abstract: Robotics and Empiricism@inproceedings{badra1998robust,
title={Robust Mosaicing Using Zernike Moments},
author={Badra, Fady and Qumsieh, Ala and Dudek, Gregory},
booktitle={Proc. Vision Interface},
pages={149--156},
year={1998},
address={Vancouver, BC},
month={June}
} abstract: This paper presents an approach to the registration of individual images to one another to produce a larger composite mosaic. The approach is based on the use of the moments of Zernike orthogonal polynomials to compute the relative scale, rotation and translation between the images. A preliminary stage involves the use of an attention-like operation to estimate potential approximate correspondence points between the images based on extrema of local edge element density. Experimental results illustrate that the technique is@inproceedings{simhon1998selecting,
title={Selecting Targets for Local Reference Frames},
author={Simhon, Saul and Dudek, Gregory},
booktitle={Proc. IEEE International Conference on Robotics and Automation},
pages={2840--2845},
year={1998},
organization={IEEE},
address={Leuven, Belgium},
publisher={IEEE Press}
} abstract: Addresses the problem of seeking out parts of the environment that provide adequate features in order to perform robot localization. The objective is to choose good regions in which local metric maps can be established. A distinctiveness measure is defined as a measure of how well the environment allows the robot to accomplish a task, in our case the task being localization. The distinctiveness measure is evaluated as a function of both the localization strategy and the environment. Areas in the environment are considered to have@inproceedings{Jenkin1998,
author = {Michael Jenkin and Evangelos Milios and Gregory Dudek},
title = {Topological Exploration of Unknown Environments with Multiple Robots},
booktitle = {Proc. of the World Automation Congress (WAC '98)},
address = {Anchorage, Alaska},
month = {May},
year = {1998},
note = {(8 pages: proceedings on CD-ROM)}
}
@incollection{Jenkin1998a,
author = {Michael Jenkin and Evangelos Milios and Gregory Dudek},
title = {Topological Exploration of Unknown Environments with Multiple Robots},
booktitle = {Robotic and Manufacturing Systems - Recent Results in Research, Development and Applications},
editor = {M. Jamshidi and F. Pierrot and M. Kamel},
volume = {7},
publisher = {TSI Press},
address = {Albuquerque, NM, USA},
year = {1998}
} abstract: Topological Exploration of Unknown Environments with Multiple Robots@inproceedings{bourque1998viewpoint,
title={Viewpoint Selection -- An Autonomous Robotic System for Virtual Environment Creation},
author={Bourque, Eric and Dudek, Gregory},
booktitle={Proceedings IEEE/RSJ Int. Conf. on Intelligent Robots and Systems (IROS)},
pages={526--531},
year={1998},
address={Victoria, BC},
month={October}
} abstract: Describes an integrated system for the automatic construction of image-based virtual realities to describe a real environment. A mobile robot autonomously navigates through the environment and uses a camera to make observations. At locations that are deemed sufficiently interesting, panoramic images are collected that are used to construct a multi-node VR movie. Images of the environment are classified in terms of two features related to human attention: edge element density and edge orientation. The system deems locations@article{oore1997mobile,
abstract = {We show how a neural network can be used to allow a mobile robot to derive an accurate estimate of its location from noisy sonar sensors and noisy motion information. The robot's model of its location is in the form of a probability distribution across a grid of possible locations. This distribution is updated using both the motion information and the predictions of a neural network that maps locations into likelihood distributions across possible sonar readings. By predicting sonar readings from locations, rather than vice versa, the robot can},
author = {Oore, Sageev and Hinton, Geoffrey E and Dudek, Gregory},
journal = {Neural Computation},
number = {3},
pages = {683--699},
pub_year = {1997},
publisher = {MIT Press One Rogers Street, Cambridge, MA 02142-1209, USA journals-info~…},
title = {A mobile robot that learns its place},
venue = {Neural Computation},
volume = {9}
} abstract: We show how a neural network can be used to allow a mobile robot to derive an accurate estimate of its location from noisy sonar sensors and noisy motion information. The robot's model of its location is in the form of a probability distribution across a grid of possible locations. This distribution is updated using both the motion information and the predictions of a neural network that maps locations into likelihood distributions across possible sonar readings. By predicting sonar readings from locations, rather than vice versa, the robot can@inproceedings{bourque1997automated,
title={Automated Creation of Image-Based Virtual Reality},
author={Bourque, Eric and Dudek, Gregory},
booktitle={Proc. SPIE Proceedings on Intelligent Systems and Manufacturing},
pages={292--303},
year={1997},
address={Pittsburgh, PA},
month={Oct}
} abstract: Furthermore, selecting suitable vantage points to produce an evocative and complete VR model is in itself an important issue. This paper deals with the automated acquisition and construction of image-based VR models by having a robotic system select and acquire images from different vantage points. The objective is to provide a fully or partially automatic SE Chen, “QuickTime VR – An image based approach to virtual environment navigation,” in Proceedings of the ACM SIGGRAPH, pp. 29–38, ACM, (New York), 1995. 16.@inproceedings{roy1997learning,
title={Learning to Rendezvous during Multi-agent Exploration or What to do When Youre Lost at the Zoo},
author={Roy, Nicholas and Dudek, Gregory},
booktitle={Proc. of the Sixth European Workshop on Learning Robots (EWLR-6)},
pages={30--45},
year={1997},
month={Aug},
address={Brighton, UK}
} abstract: We consider the problem of rendezvous between two robots collaborating in learning the layout of an unknown environment. That is, how can two autonomous exploring agents that cannot communicate with one another over long distances meet if they start exploring at different locations in an unknown environment. The intended application is collaborative map exploration. Ours is the first work to formalize the characteristics of the rendezvous problem, and we approach it by proposing several alternative algorithms that the robots@article{dudek1997map,
abstract = {This paper deals with the validation of topological maps of an environment by an active agent (such as a mobile robot), and the localization of an agent in a given map. The agent is assumed to have neither compass nor other instruments for measuring orientation or distance, and, therefore, no associated metrics. The topological maps considered are similar to conventional graphs. The robot is assumed to have enough sensory capability to traverse graph edges autonomously, recognize when it has reached a vertex, and enumerate edges},
author = {Dudek, Gregory and Jenkin, Michael and Milios, Evangelos and Wilkes, David},
journal = {Robotics and autonomous systems},
number = {2},
pages = {159--178},
pub_year = {1997},
publisher = {Elsevier},
title = {Map validation and robot self-location in a graph-like world},
venue = {Robotics and autonomous systems},
volume = {22}
} abstract: This paper deals with the validation of topological maps of an environment by an active agent (such as a mobile robot), and the localization of an agent in a given map. The agent is assumed to have neither compass nor other instruments for measuring orientation or distance, and, therefore, no associated metrics. The topological maps considered are similar to conventional graphs. The robot is assumed to have enough sensory capability to traverse graph edges autonomously, recognize when it has reached a vertex, and enumerate edges@inproceedings{Rekleitis1997,
author = {Ioannis Rekleitis and Evangelos Milios and Gregory Dudek},
title = {Multi-Robot Exploration of an Unknown Environment: Efficiently Reducing the Odometry Error},
booktitle = {Proc. International Joint Conference on Artificial Intelligence (IJCAI)},
year = {1997},
pages = {1340--1345},
address = {Nagoya, Japan},
month = sep
} abstract: This paper deals with the intelligent exploration of an unknown environment by autonomous robots. In particular, we present an algorithm and associated analysis for collaborative exploration using two mobile robots. Our approach is based on robots with range sensors limited by distance. By appropriate behavioural strategies, we show that odometry (motion) errors that would normally present problems for mapping can be severely reduced. Our analysis includes polynomial complexity bounds and a discussion of possible heuristics. 1@inproceedings{jenkin1997building,
author = {Michael Jenkin and Evangelos Milios and David Wilkes and Gregory Dudek},
title = {On building and navigating with a globally topological but locally metric map},
booktitle = {Proc. 3rd ECPD Int. Conf. on Advanced Robotics, Intelligent Automation and Active Systems},
year = {1997},
pages = {132--144},
address = {Bremen, Germany}
} abstract: On building and navigating with a globally topological but locally metric map@inproceedings{lacroix1997identification,
author = {Simon Lacroix and Gregory Dudek},
title = {On the Identification of Sonar Features},
booktitle = {Proceedings IEEE/RSJ Int. Conf. on Intelligent Robots and Systems (IROS)},
year = {1997},
pages = {586--592},
address = {Grenoble, France}
} abstract: We are interested in inferring the sources of various types of sonar features typically observed by a mobile robot. After a brief discussion of terrestrial sonar sensing, we develop a set of operators that associates arc-shaped features extracted from sonar scans with real world primitives. Our classification scheme is probabilistic and is based on empirical data: the confidence of the association hypotheses produced by the operators is evaluated statistically. Some of our experimental results suggest that methods based on models of@inproceedings{roy1997online,
title={On-line rendezvous selection for robot exploration},
author={Roy, Nicholas and Dudek, Gregory},
booktitle={Proc. American Association for Artificial Intelligence (AAAI) Workshop on On-Line Search},
pages={22--29},
year={1997},
address={Providence, RI},
note={Also available as AAAI Report WS-97-10}
} abstract: On-line rendezvous selection for robot exploration@inproceedings{Rekleitis1997,
author = {Ioannis Rekleitis and Evangelos Milios and Gregory Dudek},
title = {Reducing odometry error through cooperating robots during the exploration of an unknown world},
booktitle = {Proc. Fifth IASTED International Conference ROBOTICS AND MANUFACTURING},
year = {1997},
month = {June},
pages = {200--208}
} abstract: We consider how to cover and map an initially unknown environment using two (or more) mobile robots. Most mobile robot systems accrue odometry error while moving, and hence need to use external sensors to recalibrate their position on an ongoing basis. Unfortunately, most sensing systems are constrained with respect to the types of environment in which they are suitable. We deal with position calibration and odometry error by using multiple robots for exploration. This allows them to use one another as landmarks. We consider how@article{Tsotsos1997,
title={Shape representation and recognition from multiscale curvature},
author={Tsotsos, John K. and Dudek, Gregory},
journal={Computer Vision, Graphics and Image Processing: Image Understanding},
volume={68},
number={2},
year={1997},
pages={170--189}
} abstract: Shape Representation and Recognition from Curvature@article{Tsotsos1997,
title={Shape representation and recognition from multiscale curvature},
author={Tsotsos, John K. and Dudek, Gregory},
journal={Computer Vision, Graphics and Image Processing: Image Understanding},
volume={68},
number={2},
year={1997},
pages={170--189}
} abstract: Shape Representation and Recognition from Curvature@inproceedings{roy1997online,
title={On-line rendezvous selection for robot exploration or What to do When You're Lost at the Zoo},
author={Roy, Nicholas and Dudek, Gregory},
booktitle={Proc. American Association for Artificial Intelligence (AAAI) Workshop on On-Line Search},
pages={22--29},
year={1997},
address={Providence, RI},
note={Also available as AAAI Report WS-97-10},
url={https://cdn.aaai.org/Workshops/1997/WS-97-10/WS97-10-004.pdf}
} abstract: On-line rendezvous selection for robot exploration or What to do When YouÕre Lost at the Zoo@inproceedings{roy1997learning,
title={Learning to Rendezvous during Multi-agent Exploration or What to do When You're Lost at the Zoo},
author={Roy, Nicholas and Dudek, Gregory},
booktitle={Proc. of the Sixth European Workshop on Learning Robots (EWLR-6)},
pages={30--45},
year={1997},
organization={AAAI},
address={Brighton, UK},
note={Also appears as the AAAI Workshop AAAI Technical Report WS-97-10},
url={https://cdn.aaai.org/Workshops/1997/WS-97-10/WS97-10-004.pdf}
} abstract: Learning to Rendezvous during Multi-agent Exploration or What to do When You're Lost at the Zoo@inproceedings{ayoung1996hybrid,
title={A Hybrid Approach to 3D Representation},
author={Ayoung-Chee, Nigel and Ferrie, Frank and Dudek, Gregory},
booktitle={Proceedings of the IEEE International Workshop on Object Representation},
pages={202--209},
year={1996},
address={Cambridge, England}
} abstract: This paper deals with generic 3D shape modelling for the purposes of object recognition. Common problems with many existing methods are that they either capture insufficient detailed structure or fail to provide sufficiently abstract descriptions (global vs. local representation). As a result, they tend have a limited field of application. The approach presented here attempts to address this problem by building a composite representation of the data in terms of a superquadric augmented with multi-scale surface models. This is@article{dudek1996taxonomy,
abstract = {A key difficulty in the design of multi-agent robotic systems is the size and complexity of the space of possible designs. In order to make principled design decisions, an understanding of the many possible system configurations is essential. To this end, we present a taxonomy that classifies multi-agent systems according to communication, computational and other capabilities. We survey existing efforts involving multi-agent systems according to their positions in the taxonomy. We also present additional results concerning multi-agent},
author = {Dudek, Gregory and Jenkin, Michael RM and Milios, Evangelos and Wilkes, David},
journal = {Autonomous Robots},
pages = {375--397},
pub_year = {1996},
publisher = {Springer},
title = {A taxonomy for multi-agent robotics},
venue = {Autonomous Robots},
volume = {3}
} abstract: A key difficulty in the design of multi-agent robotic systems is the size and complexity of the space of possible designs. In order to make principled design decisions, an understanding of the many possible system configurations is essential. To this end, we present a taxonomy that classifies multi-agent systems according to communication, computational and other capabilities. We survey existing efforts involving multi-agent systems according to their positions in the taxonomy. We also present additional results concerning multi-agent@inproceedings{Ayoung-Chee1996,
author = {Nigel Ayoung-Chee and Frank Ferrie and Gregory Dudek},
title = {Enhanced 3D Representation Using a Hybrid Model},
booktitle = {Proc. International Conf. on Pattern Recognition},
year = {1996},
address = {Vienna, Austria},
month = {August},
pages = {575--579}
} abstract: This paper deals with generic 3D shape modelling for the purposes of object recognition. Difficulties with many existing methods are that they either capture insufficient detailed structure or fail to provide sufficiently abstract descriptions. The approach presented here attempts to address this problem by building a composite representation of the data in terms of a superquadric augmented with multi-scale surface models. This is illustrated experimentally using laser range data. The superquadric that results in the best possible fit@inproceedings{Ayoung-Chee1996,
author = {Nigel Ayoung-Chee and Frank Ferrie and Gregory Dudek},
title = {Enhanced 3D Representation Using a Hybrid Model},
booktitle = {Proc. International Conf. on Pattern Recognition},
year = {1996},
address = {Vienna, Austria},
month = {August},
pages = {575--579}
} abstract: This paper deals with generic 3D shape modelling for the purposes of object recognition. Difficulties with many existing methods are that they either capture insufficient detailed structure or fail to provide sufficiently abstract descriptions. The approach presented here attempts to address this problem by building a composite representation of the data in terms of a superquadric augmented with multi-scale surface models. This is illustrated experimentally using laser range data. The superquadric that results in the best possible fit@inproceedings{freedman1996environment,
title={Environment Mapping Using ``Just-In-Time'' Sensor Fusion},
author={Freedman, Paul and Rekleitis, Ioannis and Dudek, Gregory},
booktitle={Proceedings of Vision Interface},
pages={96--103},
year={1996},
address={Toronto, ON},
month={April}
} abstract: Environment Mapping Using ``Just-In-Time@article{dudek1996environment,
title={Environment mapping using multiple abstraction levels},
author={Dudek, Gregory},
journal={Proceedings of the IEEE},
volume={84},
number={11},
pages={1684--1704},
year={1996},
month={Nov},
note={Special issue on ``Signals and Symbols''}
} abstract: Environment mapping using multiple abstraction levels@article{arguin1996shape,
abstract = {A series of experiments was conducted on a patient (ELM) with bilateral inferior temporal lobe damage and category-specific visual agnosia in order to specify the nature of his functional impairment. In Experiment 1, ELM performed a task of picture/word matching that used line drawings of fruits and vegetables as stimuli. The pattern of confusions exhibited by the patient suggested a failure in processing the full range of shape features necessary for the unique specification of the target relative to other structurally related items. This},
author = {Arguin, Martin},
journal = {Visual cognition},
number = {3},
pages = {221--276},
pub_year = {1996},
publisher = {Taylor \& Francis},
title = {Shape integration for visual object recognition and its implication in category-specific visual agnosia},
venue = {Visual cognition},
volume = {3}
} abstract: A series of experiments was conducted on a patient (ELM) with bilateral inferior temporal lobe damage and category-specific visual agnosia in order to specify the nature of his functional impairment. In Experiment 1, ELM performed a task of picture/word matching that used line drawings of fruits and vegetables as stimuli. The pattern of confusions exhibited by the patient suggested a failure in processing the full range of shape features necessary for the unique specification of the target relative to other structurally related items. This@inproceedings{dudek1996just,
author = {Gregory Dudek and Paul Freedman and Yiannis Rekleitis},
title = {Just-in-Time Sensing: Efficiently Combining Sonar and Laser Range Data for Exploring Unknown Worlds},
booktitle = {Proceedings IEEE International Conference on Robotics and Automation},
year = {1996},
volume = {1},
pages = {667--672},
address = {Minneapolis, MN},
month = apr
} abstract: This paper describes an approach to combining range data from both a set of sonar sensors as well as from a directional laser range finder to efficiently take advantage of the characteristics of both types of devices when exploring and mapping unknown worlds. The authors call their approach" just in time sensing" because it uses the more accurate but constrained laser range sensor only as needed, based upon a preliminary interpretation of sonar data. In this respect, it resembles" just in time" inventory control which attempts to@inproceedings{roy1996mobile,
title={Mobile Robot Navigation: A Case Study},
author={Roy, Nicholas and Daum, Michael and Dudek, Gregory},
booktitle={Proc. American Assoc. for Artificial Intelligence National Conf. on Artificial Intelligence (AAAI)},
year={1996}
} abstract: Mobile Robot Navigation: A Case Study@inproceedings{daum1996shape,
author = {Michael Daum and Gregory Dudek},
title = {Shape from Darkness in Three Dimensions Using Information from Solar Trajectories},
booktitle = {Proc. US/Japan Student Forum (in affiliation with RSJ/IEEE IROS conference)},
year = {1996},
address = {Nagoya, Japan},
pages = {20}
} abstract: Shape from Darkness in Three Dimensions Using Information from Solar Trajectories@inproceedings{roy1996surface,
title={Surface Sensing and Classification for Efficient Mobile Robot Navigation},
author={Roy, Nicholas and Freedman, Paul and Dudek, Gregory},
booktitle={Proceedings IEEE International Conference on Robotics and Automation},
volume={2},
pages={1224--1228},
year={1996},
address={Minneapolis, MN},
month={April}
} abstract: Mobile robot navigation and localization is frequently aided by, or even dependent upon, a good estimate of the rate of dead-reckoning error accumulation. Sensor data can be used for position estimation, but this often involves overheads in acquiring and processing the data. By sensing and then classifying the surface type, an estimate of the rate of error accumulation for dead-reckoning allows one to estimate accurately how often localization, including sensor data acquisition, must be performed. The authors describe experiments in@article{dudek1996using,
abstract = {},
author = {Dudek, Gregory and Freedman, Paul and Hadjres, Souad},
journal = {Journal of Robotic Systems},
number = {8},
pages = {539--559},
pub_year = {1996},
title = {Using multiple models for environmental mapping},
venue = {NA},
volume = {13}
} abstract: @inproceedings{zhang1996vision,
title={Vision-based Robot Localization Without Explicit Object Models},
author={Zhang, Chi and Dudek, Gregory},
booktitle={Proceedings IEEE International Conference on Robotics and Automation},
volume={1},
pages={76--82},
year={1996},
organization={IEEE},
address={Minneapolis, MN},
month={April}
} abstract: We consider the problem of locating a robot in an initially-unfamiliar environment from visual input. The robot is not given a map of the environment, but it does have access to a collection of training examples, each of which specifies the video image observed when the robot is at a particular location and orientation. We address two variants of this problem: how to estimate translation of a moving robot assuming the orientation is known, and how to estimate translation and orientation for a mobile robot. Performing scene reconstruction to@inproceedings{jenkin1995experiments,
author = {Michael Jenkin and Evangelos Milios and David Wilkes and Gregory Dudek},
title = {Experiments in sensing and communication for robot convoy navigation},
booktitle = {Proceedings IEEE/RSJ Int. Conf. on Intelligent Robots and Systems (IROS)},
year = {1995},
volume = {2},
pages = {268--273},
address = {Pittsburgh, PA},
month = {August}
} abstract: This paper deals with coordinating behaviour in a multi-autonomous robot system. When two or more autonomous robots must interact in order to accomplish some common goal, communication between the robots is essential. Different inter-robot communications strategies give rise to different overall system performance and reliability. After a brief consideration of some theoretical approaches to multiple robot collections, we present concrete implementations of different strategies for convoy-like behaviour. The convoy@inproceedings{jenkin1995exploring,
title={Exploring Graph-Like World Embedded in a Metric Map},
author={Jenkin, Michael and Milios, Evangelos and Wilkes, David and Dudek, Gregory},
booktitle={Proc. Vision Interface},
pages={195--202},
year={1995},
address={Quebec City, Que.},
month={July}
} abstract: Exploring Graph-Like World Embedded in a Metric Map@inproceedings{romanik1995localizing,
title={Localizing a Robot with Minimum Travel},
author={Romanik, Kathleen and Whitesides, Sue and Dudek, Gregory},
booktitle={SIAM Symposium on Discrete Algorithms (SODA)},
pages={437--446},
year={1995},
month={Jan}
} abstract: We consider the problem of localizing a robot in a known environment modeled by a simple polygon P. We assume that the robot has a map of P but is placed at an unknown location inside P. From its initial location, the robot sees a set of points called the visibility polygon V of its location. In general, sensing at a single point will not suffice to uniquely localize the robot, since the set H of points in P with visibility polygon V may have more than one element. Hence, the robot must move around and use range sensing and a compass to@inproceedings{jenkin1995multi,
author = {Michael Jenkin and Evangelos Milios and David Wilkes and Gregory Dudek},
title = {Multi-robot landmark-based self-location and exploration},
booktitle = {Proc. Third Int. Symposium on Intelligent Robotic Systems (SIRS)},
year = {1995},
pages = {49--56},
address = {Pisa, Italy},
month = {July}
} abstract: Multi-robot landmark-based self-location and exploration@inproceedings{langer1995space,
title={Space occupancy using multiple shadowimages},
author={Langer, Michael and Zucker, Steven W. and Dudek, Gregory},
booktitle={Proceedings IEEE/RSJ Int. Conf. on Intelligent Robots and Systems (IROS)},
volume={1},
pages={285--290},
year={1995},
address={Pittsburgh, PA},
month={August}
} abstract: Addresses the problem of estimating 3D space occupancy using video imagery in the context of mobile robotics. A stationary robot observes a cluttered scene from a single viewpoint, and a second robot illuminates the scene from a sequence of directions thus producing a sequence of grey-level images. Differences of successive images are used to compute a sequence of shadowimages. The problem is to compute free space and occupied space from these shadowimages. Solutions to this problem are known for the special case of@inproceedings{pateras1995understanding,
title={Understanding Referring Expressions in a Person-Machine Spoken Dialogue},
author={Pateras, Claudia and DeMori, Renato and Dudek, Gregory},
booktitle={Proc. of the IEEE Conference of Acoustics, Speech and Signal Processing},
pages={197--200},
year={1995},
address={Detroit, MI},
month={April}
} abstract: In the domain of mobile robotic task execution under dialogue control, a primary goal is to identify the task target which is specified by a natural language description. A number of concepts are expressed in the user spoken language by vague terms like" the big box" and" very close to the door". We use fuzzy logic to map these vague terms onto the quantitative data collected by system sensors. Fuzziness may cause uncertainty in interpretation and, in particular, in understanding references. This uncertainty is abated by collecting additional@article{Arguin1994,
author = {Martin Arguin and Daniel Bub and Gregory Dudek},
title = {Dimensional Decomposition of Human Shape Recognition},
journal = {Investigative Ophthalmology and Visual Science (Suppl)},
year = {1994},
publisher = {The Association for Research in Vision and Opthalmology},
address = {Sarasota, FL},
month = {May}
} abstract: Dimensional Decomposition of Human Shape Recognition@inproceedings{jenkin1994horoptor,
author = {Michael Jenkin and J. K. Tsotsos and Gregory Dudek},
title = {Horoptor and active cyclotorsion},
booktitle = {Proceedings of the International Conference on Pattern Recognition},
year = {1994},
volume = {1},
pages = {707--710}
} abstract: When a particular 3D point is fixated by a robotic stereo system different portions of the world are brought into interocular alignment. This region is known as the horoptor. Purposeful modifications to the binocular geometry can be used to bring different regions of three-space closer to the horoptor: camera pan and tilt define the rough structure of the horoptor, while camera torsion can be used to change its local shape. Theoretical and empirical results suggest that for binocular vision tasks: 1) it is important to understand the region of three@inproceedings{Arguin1994,
author = {Martin Arguin and Daniel Bub and Gregory Dudek},
title = {Human Integration of Shape Primitives},
booktitle = {Proceedings of the International Workshop on Visual Form},
year = {1994},
pages = {130--138},
address = {Capri, Italy},
month = {May}
} abstract: This paper deals with human shape recognition. In particular, we present results relating the role of human inferior temporal cortex to the description and recognition of shapes using a parametric three-dimensional shape space. In experiments with a patient with damaged IT cortex, we show that his inability to recognize or distinguish between members of a large family of simple shapes can be traced to an inability to simultaneously combine information related to multiple global shape dimensions. We have discovered that the relevant@inproceedings{freedman1994mapping,
title={Mapping Unknown Graph-Like Worlds},
author={Freedman, Paul and Dudek, Gregory},
booktitle={Proceedings of the International Advanced Robotics Programme Workshop on Robotics in Space},
pages={1--20},
year={1994},
address={Montreal, Canada},
month={July}
} abstract: We consider the problem of constructing a map of an unknown environment by an autonomous agent such as a mobile robot. Because accurate positional information is often difficult to ensure, we consider the problem of exploration in the absence of metric (positional) information. Worlds are represented by graphs (not necessarily planar) consisting of a fixed number of discrete places linked by bidirectional paths. We assume the robot can consistently enumerate the edges leaving a vertex (that is, it can assign a cyclic@inproceedings{alami1994multi,
title={Multi-scale object representation using surface patches},
author={Alami, Wassim and Dudek, Gregory},
booktitle={Proceedings of the International Society for Optical Engineering},
volume={2353},
pages={108--119},
year={1994}
} abstract: We introduce an approach to the representation of curved or polyhedral 3-D objects and apply this representation to pose estimation. The representation is based on surface patches with uniform curvature properties extracted at multiple scales. These patches are computed using multiple alternative decompositions of the surface based on the signs of the mean and Gaussian curvatures. Initial coarse decompositions are subsequently refined using a curvature compatibility scheme to rectify the effect of noise and quantization errors. The@inproceedings{zhang1994pose,
title={Pose Estimation From Image Data Without Explicit Object Models},
author={Zhang, Chi and Dudek, Gregory},
booktitle={Proceedings of Vision Interface},
year={1994},
month={May},
address={Banff, Alta.}
} abstract: We consider the problem of locating a robot in an initially-unfamiliar environment from visual input. The robot is not given a map of the environment, but it does have access to a limited set of training examples each of which specifies the video image observed when the robot is at a particular location and orientation. Such data might be acquired using dead reckoning the first time the robot entered an unfamiliar region (using some simple mechanism such as sonar to avoid collisions). In this paper, we address a specific variant of this problem for@inproceedings{MacKenzie1994,
author = {Paul MacKenzie and Gregory Dudek},
title = {Precise Positioning Using Model-Based Maps},
booktitle = {Proceedings of the 1994 IEEE International Conference on Robotics and Automation},
year = {1994},
pages = {1615--1621},
address = {San Diego, CA},
month = {May}
} abstract: This paper addresses the coupled tasks of constructing a spatial representation of the environment with a mobile robot using noisy sensors (sonar) and using such a map to determine the robot's position. The map is not meant to represent the actual spatial structure of the environment so much as it is meant to represent the major structural components of what the robot" sees". This can, in turn, be used to construct a model of the physical objects in the environment. One problem with such an approach is that maintaining an absolute@inproceedings{jenkin1993multi,
author = {Michael Jenkin and Gregory Dudek},
title = {A Multi-Layer Distributed Environment for Mobile Robots},
booktitle = {Proceedings of the International Conference on Intelligent Autonomous Systems: IAS-3},
year = {1993},
address = {Pittsburgh, PA},
month = feb,
pages = {542--550}
} abstract: A Multi-Layer Distributed Environment for Mobile Robots@inproceedings{jenkin1993taxonomy,
author = {Michael Jenkin and Evangelos Milios and David Wilkes and Gregory Dudek},
title = {A Taxonomy for swarm robotics},
booktitle = {Proceedings IEEE/RSJ Int. Conf. on Intelligent Robots and Systems (IROS)},
year = {1993},
pages = {441--447},
address = {Yokohama, Japan},
month = {July}
} abstract: A Taxonomy for swarm robotics@inproceedings{jenkin1993map,
title={Map validation and self-location in a graph-like world},
author={Jenkin, Michael and Milios, Evangelos and Wilkes, David and Dudek, Gregory},
booktitle={Proceedings of the International Joint Conference of Artificial Intelligence (IJCAI-93)},
pages={1648--1653},
year={1993},
address={Chambery, France},
month={August}
} abstract: We present algorithms for the discovery and use of topological maps of an environment by an active agent (such as a person or a mobile robot). We discuss several issues dealing with the use of pre-existing topological maps of graph-like worlds by an autonomous robot and present algorithms, worst cases complexity, and experimental results (for representative real-world examples) for two key problems. The first of these problems is to verify that a given input map is a correct description of the world (the validation problem). The second is to@inproceedings{MacKenzie1993,
author = {Paul MacKenzie and Gregory Dudek},
title = {Model Based Map Construction for Mobile Robot Localization},
booktitle = {Proceedings of Vision Interface '93},
year = {1993},
address = {Toronto, Ontario},
month = {July},
pages = {97--102}
} abstract: Model Based Map Construction for Mobile Robot Localization@inproceedings{jenkin1993multi,
author = {Michael Jenkin and Evangelos Milios and David Wilkes and Gregory Dudek},
title = {Multi-Transducer Sonar Interpretation},
booktitle = {Proceedings of the 1993 IEEE International Conference on Robotics and Automation},
year = {1993},
pages = {392--397},
address = {Atlanta, GA},
month = {May}
} abstract: In response to difficulties with interpretation of data from narrow beam time-of-flight sonar for robotics, an algorithm for sonar interpretation that uses the entire return signals from several transducers with broad, overlapping fields of view is presented. The result is an algorithm that reconstructs the geometry in front of the robot with only a moderate amount of robot motion. Preliminary results with a three transducer system are shown. The results demonstrate an unusual ability to recover the presence of obstacles even when they are@inproceedings{jenkin1993utility,
author = {Michael Jenkin and Evangelos Milios and David Wilkes and Gregory Dudek},
title = {On the Utility of Multi-Agent Autonomous Robot Systems},
booktitle = {Proceedings of the International Joint Conference of Artificial Intelligence (IJCAI-93) Workshop on Dynamically Interacting Robots},
year = {1993},
address = {Chambery, France},
pages = {101--108}
} abstract: @inproceedings{dudek1993organizational,
author = {Gregory Dudek and Michael Jenkin and Evangelos Milios and David Wilkes},
title = {Organizational Characteristics for Multi-Agent Robotic Systems},
booktitle = {Proceedings of Vision Interface '93},
year = {1993},
address = {Toronto, Ontario},
month = {July},
pages = {91--96}
} abstract: @inproceedings{jenkin1993robust,
author = {Michael Jenkin and Evangelos Milios and David Wilkes and Gregory Dudek},
title = {Robust Positioning with a Multi-Agent Robotic System},
booktitle = {Proceedings of the International Joint Conference of Artificial Intelligence (IJCAI-93) Workshop on Dynamically Interacting Robots},
year = {1993},
pages = {118--123},
address = {Chambery, France},
month = {August}
} abstract: A collection of interacting autonomous robots can de ne a local coordinate system with respect to one another without reference to environmental features. This simpli es tasks requiring robots to occupy or traverse a set of positions in the environment, such as mapping, conveyance and search. We argue for an approach to positioning in which sensing errors remain localized, and dead-reckoning plays no role. This involves a robot-based representation for the environment, in which metric information is used locally to@inproceedings{Hadjres1993,
author = {Souad Hadjres and Paul Freedman and Gregory Dudek},
title = {Using local information in a non-local way for mapping graph-like worlds},
booktitle = {Proceedings of the International Joint Conference of Artificial Intelligence (IJCAI-93)},
year = {1993},
pages = {1639--1645},
address = {Chambery, France},
month = {August}
} abstract: This paper describes a technique whereby an autonomous agent such as a mobile robot can explore an unknown environment and make a topological map of it. It is assumed that the environment can be represented as a graph, that is, as a xed set of discrete locations or regions with an ordered set of paths between them. In previous work, it has been shown that such worlds can be fully explored and described using a single movable marker even if there are no spatial metrics and almost no sensory ability on the part of the robot. Here we@techreport{Dudek1993,
author = {Gregory Dudek and Michael Jenkin and Evangelos Milios and David Wilkes},
title = {Reflections on Sonar Range Sensing},
institution = {CIM},
number = {CIM-92-9},
year = {1993},
month = {December 21}
} abstract: Reflections on Sonar Range Sensing@inproceedings{Hadjres1992,
author = {Souad Hadjres and Paul Freedman and Gregory Dudek},
title = {Algorithms for Active Exploration of Unknown Environments: Using Uncertain Sensing Data to Create a Reliable Map},
booktitle = {Proceedings of the International Society for Optical Engineering Symposium on Advances in Intelligent Robotics Systems: Conference on Mobile Robotics VII},
address = {Boston, MA},
month = {November},
year = {1992}
} abstract: Algorithms for Active Exploration of Unknown Environments: Using Uncertain Sensing Data to Create a Reliable Map@incollection{jenkin1992modelling,
author = {Michael Jenkin and Evangelos Milios and David Wilkes and Gregory Dudek},
title = {Modelling Sonar Range Sensors},
booktitle = {Advances in Machine Vision: Strategies and Applications},
editor = {C. Archibald and E. Petriu},
publisher = {World Scientific Press},
address = {Singapore},
year = {1992},
pages = {361--370}
} abstract: Modelling Sonar Range Sensors@inproceedings{dudek1992robot,
author = {Gregory Dudek},
title = {Robot Map-Making Using Weak Sensory Feedback},
booktitle = {Proceedings of the Workshop on Sensor-Based Mobile Robotics},
address = {Nice, France},
month = {May},
year = {1992}
} abstract: Robot Map-Making Using Weak Sensory Feedback@article{dudek1992shape,
author = {Gregory Dudek},
title = {Shape Classification and Scale-Space Texture},
journal = {Investigative Ophthalmology and Visual Science (Suppl)},
year = {1992},
publisher = {The Association for Research in Vision and Ophthalmology},
address = {Sarasota, FL},
month = {May}
} abstract: @inproceedings{dudek1992shape,
author = {Gregory Dudek},
title = {Shape Description and Classification using Scale-Space Measurement},
booktitle = {Proceedings of the Workshop on Shape in Picture},
year = {1992},
address = {Driebergen, Holland},
month = {August},
note = {Revised manuscript to appear in the book Shape in Picture published by Springer Verlag}
} abstract: Shape Description and Classification using Scale-Space Measurement@incollection{tsotsos1992using,
author = {John K. Tsotsos and Gregory Dudek},
title = {Using Curvature Information in the Decomposition and Representation of Planar Curves},
booktitle = {Active Perception and Robot Vision},
editor = {A. Sood and H. Wechsler},
publisher = {Springer-Verlag},
year = {1992},
pages = {527--536}
} abstract: This paper describes a new symbolic representation for planar curves. This representation is based on a segmentation of the curve based on regions of uniform curvature. Rather than smooth noisy data before doing the decomposition, the technique defines a family of functions that extract the segments of the curve as part of the smoothing process. The representation decomposes the curve at multiple scales and the parts produced appear to correspond to a natural decomposition of the curve. It also allows for multiple descriptions of@article{dudek1978robotic,
abstract = {We address the problem of robotic exploration of a graphlike world, where no distance or orientation metric is assumed of the world. The robot is assumed to be able to autonomously traverse graph edges, recognize when it has reached a vertex, and enumerate edges incident upon the current vertex relative to the edge via which it entered the current vertex. The robot cannot measure distances, and it does not have a compass. We demonstrate that this exploration problem is unsolvable in general without markers, and, to solve it, we equip},
author = {Dudek, Gregory and Jenkin, Michael and Milios, Evangelos and Wilkes, David},
journal = {J. Comput., vol},
number = {3},
pub_year = {1978},
title = {Robotic exploration as graph construction},
venue = {J. Comput., vol},
volume = {7}
} abstract: We address the problem of robotic exploration of a graphlike world, where no distance or orientation metric is assumed of the world. The robot is assumed to be able to autonomously traverse graph edges, recognize when it has reached a vertex, and enumerate edges incident upon the current vertex relative to the edge via which it entered the current vertex. The robot cannot measure distances, and it does not have a compass. We demonstrate that this exploration problem is unsolvable in general without markers, and, to solve it, we equip@inproceedings{tsotsos1991robustly,
title={Robustly recognizing curves using curvature-tuned smoothing},
author={Tsotsos, John K. and Dudek, Gregory},
booktitle={1991 IEEE Conference on Computer Vision and Pattern Recognition},
pages={35--41},
year={1991},
organization={IEEE},
address={Maui, HI},
month={July}
} abstract: Robustly recognizing curves using curvature-tuned smoothing@inproceedings{dudek1991shape,
author = {Gregory Dudek},
title = {Shape Metrics From Curvature Scale-Space and Curvature-Tuned Smoothing},
booktitle = {Proceedings of the Conference on Geometric Methods in Computer Vision},
address = {San Diego, CA},
month = {July},
year = {1991}
} abstract: This research deals with the decomposition and description of curved objects. In ongoing work, a new part description for curves and surfaces using a set of curvature-based minimization operators has been developed. The decomposition operation simultaneously performs data interpolation, data smoothing, and segmentation. The unification of these three stages results in a smoothing operation that is tightly coupled with the primitives to be used in subsequent object description. Each of the minimization operators, in addition to@inproceedings{jenkin1991simulation,
author = {Michael Jenkin and Evangelos Milios and David Wilkes and Gregory Dudek},
title = {The Simulation of Sonar Mapping in Complex Environments Using Multiple Reflecting Surfaces},
booktitle = {Proceedings of Vision Interface '91},
year = {1991},
pages = {213--217},
address = {Calgary, Alta.},
month = {July}
} abstract: @inproceedings{tsotsos1990goal,
title={Goal-directed Smoothing for the Curvature-based Segmentation of 3-Dimensional Surfaces},
author={Tsotsos, John K. and Dudek, Gregory},
booktitle={Proceedings of the Canadian Society for the Computational Studies of Intelligence},
pages={253--257},
year={1990},
address={Ottawa, Ontario},
month={May}
} abstract: Gregory Dudek Goal-directed smoothing for the curvature-based segmentation of 3-dimensional surfaces Pages 253–257@inproceedings{dudek1990object,
author = {Gregory Dudek},
title = {Object Description Using Qualitative Surface Descriptors},
booktitle = {Proceedings of the AAAI-90 Workshop on Qualitative Vision},
address = {Boston, MA},
month = {July},
year = {1990}
} abstract: Object Description Using Qualitative Surface Descriptors@inproceedings{tsotsos1990recognizing,
title={Recognizing Planar Curves Using Curvature-Tuned Smoothing},
author={Tsotsos, John K. and Dudek, Gregory},
booktitle={Proceedings of the 10th International Conference of Pattern Recognition},
pages={130--135},
year={1990},
month={June},
address={Atlantic City, N.J.}
} abstract: uses curvature information and tracks the derivative of curvature across scales to produce a syntactic description of curves. The In this paper the authors present a new technique for both the fact that explicit parts are extracted by both these approaches smoothing and decomposition of planar curves. This technique, is a very useful and appealing characteristic. Contour informadubbed “curvature-tuned smoothing” provides for robust rota-tion has traditionally been easier to extract from images than tion and translation@inproceedings{tsotsos1990decomposition,
author = {John K. Tsotsos and Gregory Dudek},
title = {The Decomposition and Representation of Planar Curves},
booktitle = {Proceedings of the Conference on Curves and Surfaces in Computer Vision and Graphics},
year = {1990},
address = {Santa Clara, CA},
month = {February}
} abstract: The Decomposition and Representation of Planar Curves@inproceedings{jenkin1990robust,
title={The Robust Simulation of Sonar Mapping from Multiple Viewpoints},
author={Jenkin, Michael and Milios, Evangelos and Wilkes, David and Dudek, Gregory},
booktitle={Proceedings of the International Society for Optical Engineering Symposium on Advances in Intelligent Robotics Systems: Conference on Mobile Robotics V},
pages={536--542},
year={1990},
address={Boston, MA},
month={November}
} abstract: The Robust Simulation of Sonar Mapping from Multiple Viewpoints@inproceedings{jenkin1989using,
author = {Michael Jenkin and Evangelos Milios and David Wilkes and Gregory Dudek},
title = {Using a Marker to Map an Unknown Environment},
booktitle = {Proceedings Vision Interface '89},
year = {1989},
pages = {143--150},
address = {London, Ontario},
month = {June}
} abstract: Using a Marker to Map an Unknown Environment@inproceedings{jenkin1989using,
title={Using Multiple Markers in Graph Exploration},
author={Jenkin, Michael and Milios, Evangelos and Wilkes, David and Dudek, Gregory},
booktitle={Proceedings of the International Society for Optical Engineering Symposium on Advances in Intelligent Robotics Systems: Conference on Mobile Robotics},
pages={77--87},
year={1989},
address={Philadelphia, PA},
month={November}
} abstract: A fundamental problem in robotics is that of exploring an unknown environment. Most current approaches to exploration make use of a global distance metric that is used to relate past sensory experiences to local measurements. Rather than rely on such an assumption we consider the more general problem of exploration without a distance metric, as is typical of exploring using only visual information: we propose robot exploration as graph building. In earlier papers we have shown that it is not possible for a robot to successfully explore a@inproceedings{jenkin1986make,
author = {Michael Jenkin and Howard Marcus and Gregory Dudek},
title = {How to Make Friends With Number-Crunchers: Adding Single-User Array-Processor Slave environment to VAX UNIX},
booktitle = {Proceedings of 1986 USENIX Conference},
year = {1986},
address = {Atlanta, GA},
month = {June},
pages = {200--208}
} abstract: How to Make Friends With Number-Crunchers: Adding Single-User Array-Processor Slave environment to VAX UNIX@inproceedings{hamacher1985design,
title={Design of a Microcomputer-Based Central File Server},
author={Hamacher, V. Carl and Dudek, Gregory and Holt, Richard C.},
booktitle={Proceedings of the CIPS Congress},
pages={176--182},
year={1985},
address={Montreal, Quebec},
month={June}
} abstract: Design of a Microcomputer-Based Central File Server@inproceedings{swanston1983overview,
author = {Edward N. Swanston and David B. Richardson and Hugh Campbell and Gregory Dudek},
title = {An Overview of the Metropolitan Toronto Traffic Control Computer System},
booktitle = {Proceedings of the 1983 International Electrical and Electronics Conference},
year = {1983},
address = {Toronto, Ontario},
month = {March},
pages = {202--208}
} abstract: An Overview of the Metropolitan Toronto Traffic Control Computer System