MRL Members |
Faculty
- Prof. Gregory Dudek: CIM Director, MRL Director, Assistant Professor (SOCS).
Prof. Dudek does research on sensing for mobile robotics including vision, robot pose estimation (position estimation), recognition, path planning. He is also looking into a few non-robotic problems such as those related to recommender systems.
homepage
e-mail
Current Students
- Philippe Giguere: Ph.D. Candidate
Second year PhD student working on the AQUA project.
homepage
e-mail
- Dimitri Marinakis: Ph.D. Candidate
I am a PhD candidate doing research related to sensor networks. Specifically, I am working on a developing a probabilistic inference algorithm that will
determine the topology of the sensors based on observational data collected independently at each node.
homepage
e-mail
- Junaed Sattar: Ph.D. Candidate
I am currently a first year PhD student working towards robot autonomy for the AQUA underwater robot project. Specifically,
I'm working on a vision-based navigational system for AQUA which will enable her to track a particular (moving) object underwater and steer
towards that target on her own. I am also investigating possibilities of combining machine learning methods with the visual tracking subsystem to come
up with a robust, accurate and stable serviong system, that in theory can be independent of the actual robot hardware.
homepage
e-mail
- Anqi Xu: Ph.D. Candidate
I am currently a first year Ph.D. student studying at McGill University's School of Computer Science. My research interests include Human-Robotic Interaction topics and Augmented Reality systems.
homepage
e-mail
- Gabriel Charette: B.Sc. Candidate
I am a second year Bachelor of Science student in Honours Computer Science. I am currently working on a summer research project which involves building a python module that will allow for command line interaction with the AQUA robot.
homepage
e-mail
- Yasmina Schoueri: M.Sc. Candidate
I am an M.Sc. candidate at the School of Computer Science. I am currently working on a recommendation system for mobile applications.
homepage
e-mail
Past Students
- Matthew Garden: Ph.D. Candidate
I am currently a PhD student working on a new approach to recommender/Collaborative Filtering
in which a user not only specifies which items s/he (dis)likes but also which features of the
item influenced that overall impression of the item. We believe that knowing this extra information
will allow us to better predict what the user will actually enjoy. Our
implementation of this idea is the Recommendz system.
homepage
e-mail
- Paul DiMarco: M.Sc. Graduate, 2006
I'm a MSc student working towards a monocular-vision-based localization
algorithm for underwater robots. The problem differs from
localization techniques for land-based robots in many ways, from the addition
of three more degrees of freedom to the elimination of useful constraints.
homepage
e-mail
- David Meger : M.Sc. Graduate, 2006
I am a M.Sc. student researching smart environments where a network of cameras (such as a security system) can aid robot navigation.
This work focusses on automating camera calibration and producing maps of the camera locations in an instance of visual Simultaneous Localization and Mapping.
homepage
e-mail
- Daniel Burfoot : M.Sc. Graduate, 2006
The goal of my work is to use facial gestures to control robots (such as a mechanized wheelchair) or software applications (such as a video game).
Additionally I plan to use context information
(eg, the location of the robot) to improve the reliability of the control. I am co-supervised by Greg Dudek and Joelle Pineau.
homepage
e-mail
- Derek Johns : M.Sc. Graduate, 2006
I am a MSc candidate working on the problem of vision-based localization in urban environments.
My approach uses horizon information from images to estimate position. This solution is attractive
because of the unreliable nature of GPS in cities where the reflection of radio signals off buildings can cause noisy pose estimation.
homepage
e-mail
- Eric Bourque : Ph.D. Graduate, 2005
Ph.D. graduate worked on procedural texturing. Specifically, automatically estimating the correct
parameters to create a perceptually similar procedural representation of a given texture sample.
homepage
e-mail
- Saul Simhon: Ph.D. Graduate, 2005
PhD graduate who worked on a Sketch understanding system. The
goal is to make it easier for users to produce graphics by allowing them to
describe a scene using only a rough, hand-drawn outline. The system would
then elaborate the outline, correcting the shape of curve strokes and filling in texture and
color to produce a refined illustration. The system learns the applicable
refinements from example so it can be applied to arbitrary scene types.
homepage
e-mail
- Luz Abril Torres-Méndez: Ph.D. Graduate, 2005
PhD graduate worked on the problem of inferring complete range maps where only video images and very limited range data is available.
The goal is to facilitate the building of 3D environment models by exploiting the fact that both video imaging and limited range sensing
are ubiquitous readily-available technologies while complete volume scanning is prohibitive on most mobile plataforms. Our methodology is
to compute the relationship between the observed range data and the variations in the intensity image, and use this to extrapolate/interpolate
new range values.
homepage
e-mail
- Robert Sim: Ph.D. Graduate
PhD graduate worked on problems in mobile robot localization
and environment learning using vision and range sensors. In short, he
examined
questions like: Given a map, and a sensor reading, what is the
most plausible place I might be in? How can I move to improve my confidence
about where I am? What are the best places to move to to build a map? What is
the best representation for a map? How can attention be used to simplify the
map-building process?
homepage
e-mail
- Malvika Rao: M.Sc. Graduate
MSc graduate developed a randomized approximation algorithm for robot localization in self-similar environments.
The robot is placed at an unknown location inside a self-similar environment modeled by a simple polygon P.
It has a map of P and can sense its environment thereby computing visibility data. However, the self-similarities
in the environment mean that the same visibility data may correspond to several different locations. Robot localization
with minimum travel in a self-similar environment was shown to be NP-hard
(DRW, ``Localizing a Robot with Minimum Travel'', SIAM J. on Computing, '98).
homepage
e-mail
- Ioannis Rekleitis: Ph.D. Graduate
Adjunct Professor of Computer Science at McGill, and employee of the Canadian Space Agency.
Worked on Cooperative Localization Mapping and Exploration; state estimation; Sensor Networks.
homepage
e-mail
- Sumedha Ahuja: M.Sc. Graduate, 2005
Sumedha worked in the field of Information Retrieval and Data Mining. Her project dealt with recommendation systems and preference elicitation.
She was co-supervised by Professors Doina Precup and Gregory Dudek. She graduated June 2005.
homepage
e-mail
- Richard Unger M.Sc. Candidate
Richard is an international student from Austria currently pursueing his M.Sc. at the Mobile Robotics Lab. Richard's research is in gesture
recognition for robot control. Specifically he uses probabalistic learning algorithms (HMMs) to train and then recognize a user's
(who is wearing a specially coloured glove) gestures from input images captured with a video camera.
homepage
e-mail
- Deeptiman Jugessur: M.Sc. Graduate, 1999
e-mail
- Scott Burlington: M.Sc. Graduate, 1999
Scott finished his work in efficient planar search algorithms at
CIM in 1999. Specifically he studied the efficacy of logarithmic spiral search patterns
for robotic search tasks, automated map discovery and autonomous navigation versus other
traditional approaches. His focus is now in applied computational geometry and various
scientific software development. He continues to work here in Montreal.
homepage
e-mail
- Michael Daum: M.Sc. Graduate, 1999
e-mail
- Nicholas Roy: M.Sc. Graduate, 1997
homepage
e-mail
- Francois Belair: M.Sc. Graduate, 1999
e-mail
Past Staff
Past Visitors
- Michel Taix:
Visiting Professor to CIM in 1999.
e-mail
|
|