Personal Data: -------------- Name: Jesse Hoey Address: Department of Computer Science, University of Toronto, 6 King's College Rd., Toronto, Ont, M5S 3H5 Phone: 1-416-946-0857 Email: jhoey@cs.toronto.edu Web: http://www.cs.toronto.edu/~jhoey/ Research Interests: ------------------- My primary interest is in value-directed video processing and analysis. My secondary interest is detection and recognition of events in video. My strengths are in visual machine learning and decision making in very large uncertain environments, including cognitive assistive technologies and service robotics. I have worked in diverse areas of computational intelligence, including machine learning, control and decision theory, computer vision, robotics and reinforcement learning. Education: --------- 1997-2004 Ph.D. Computer Science, University of British Columbia, Vancouver, Canada 1992-1994 M.Sc. Physics, University of British Columbia, Vancouver, Canada 1989-1992 B.Sc. Physics (Honours), McGill University, Montreal, Canada Research Experience: -------------------- 2004-Present - Postdoctoral Fellow, Departments of Occupational Therapy and Computer Science (cross-appointed), University of Toronto, Supervisors: Dr. Alex Mihailidis and Dr. Craig Boutilier . Working on the development of intelligent supportive environments for adults. Developing computer vision and decision theoretic models of human behaviors for use by an intelligent system for helping people with cognitive disabilities perform daily living tasks. Researching approximate planning methods for partially observable Markov decision processes (POMDPs). Developing a wheelchair collision avoidance system using 3D sensors. Developing methods for detecting and preventing unsafe stair use in older adults. Aiding in the development of fall detection and human activity monitoring systems using computer vision. 2000-2004 Research Assistant, Department of Computer Science ,UBC Supervisor: Dr. James J. Little. Primary project: computer vision analysis of human motion. Developed novel representations of motion in the human face. Built a software system for tracking and analysing human facial expressions in natural interactions. Secondary project: mobile robotics. Developed software for mobile robots including the lab's robotic waiter, Jose, and the robotic messenger, HOMER. Developed software for control of IEEE 1394 (firewire) single-lens and stereo digital video cameras. Participated in multiple conference and in-house robotic demonstrations. Assisted junior graduate students in the laboratory. Wrote conference papers and technical reports. 2002-2003 Research Associate, Computer Vision Driver Analysis project Nissan Motor Corp. -- UBC partnership. Project supervised by Dr. James J. Little and Dr. Nando de Freitas. Principal researcher in project to analyse the motions of car drivers. Analysed videos of drivers to find correlations between driver motions and driver workload, car speed, and traffic density. Wrote research proposals, interim reports and a final report. Gave presentations and demonstrations to senior Nissan executives and researchers, in Canada and in Japan. Awards: ------- 2005 Canesta (TM) Vision Contest Grand Prize Winner Innovative Application of Canesta's electronic perception technology, Wheelchair collision obstacle avoidance (\$17,500 USD Cash and in-kind prize) 2001 First place, 2001 {\em Hors D'\oe uvres Anyone?} Mobile Robot Competition Team member, Seattle, WA, August, 2001. 1998,1999,2000 Teaching Assistant Award, Department Computer Science, UBC Awarded to departmental teaching assistants for outstanding student evaluations and faculty recommendations. 1992 Horace Watson Medal. McGill University, Montreal, Canada. Awarded for highest academic standing in Honours Physics Professional Activities: ------------------------ Reviewer for major computer science journals: PAMI 2004,2005; CVIU 2004; IMAVIS 2004. Reviewer for major international conferences in computer vision, computer graphics and artificial intelligence: IJCAI 2005; UAI 2005; RSS 2005; MICAI 2005; GI 2004; CVPR 2001,2003,2005; ICPR 2001; ECCV 2002; EG 2002; KI 2002 Wrote SPUDD, an efficient planning engine based on Markov decision Processes (MDPs) and algebraic decision diagrams (ADDs). See www.cs.toronto.edu/~jhoey/Spudd/ Research Grants: ---------------- 2006 - Precarn-CITO (Communications and Information Technology Ontario) co-Investigator: Intelligent Haptic Stroke Rehabilitation amount: $599,990(CDN) 2005 - Canadian Institute for Health Research co-Investigator: Development and validation of an automated tool for detecting and preventing unsafe stair use by older adults. amount: $48,227(CDN) 2002 - Nissan Motor Corp., Japan -- UBC Computer Science Partnership Grant Analysis of driver behavior with computer vision amount: $95,275(CDN) Research Personel Trained or Supervised: ---------------------------------------- Dan Gunn, Research Assistant, IATSL 05/2005-09/2005 The development of an intelligent anti-collision system for a powered wheelchair. Jasper Snoek, Research Assistant, IATSL 05/2005-09/2005 Development of an automated tool for detecting unsafe stair use by older adults Chris Riddle, Department of Occupational Therapy, University of Toronto 04/2005-06/2005 Clinical testing of an automated prompting system for handwashing Publications: ------------- 1. Refereed Journal Articles Jesse Hoey and James J. Little. Value-Directed Human Behavior Analysis with Partially Observable Markov Decision Processes. Accepted for publication in IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI), 2006. Jen Boger, Jesse Hoey, Pascal Poupart, Craig Boutilier, Geoff Fernie, and Alex Mihailidis. A planning system based on Markov decision processes to guide people with dementia through activities of daily living. Accepted with revisions to IEEE Transactions on Information Technology in Biomedecine, 2005 2. Refereed Conference Articles Jesse Hoey and Pascal Poupart. Solving POMDPs with Continuous or Large Discrete Observation Spaces. In Proc. of International Joint Conference on Artificial Intelligence (IJCAI), Edinburgh, Scotland, July 2005 Jen Boger, Pascal Poupart, Jesse Hoey, Craig Boutilier, Geoff Fernie, and Alex Mihailidis. A Decision-Theoretic Approach to Task Assistance for Persons with Dementia. In Proc. of International Joint Conference on Artificial Intelligence (IJCAI), Edinburgh, Scotland, July 2005 Jesse Hoey, Pascal Poupart, Craig Boutilier and Alex Mihailidis. Semi-Supervised Learning of a POMDP model of Patient-Caregiver Interactions. In Proc. IJCAI Workshop on Modeling others from Observations (MOO), Edinburgh, Scotland, July 2005 Jesse Hoey and James J. Little. Value Directed Learning of Facial Displays. In Proc of IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Washington, DC, June 2004 Jesse Hoey and James J. Little. Decision Theoretic Modeling of Human Facial Displays. In Proc. of 8th European Conference on Computer Vision (ECCV), Prague, CZ, May 2004 Pantelis Elinas, Enrique Sucar, Alberto Reyes and Jesse Hoey. A Decision Theoretic Approach for Task Coordination in Social Robots. In Proc. of IEEE International Workshop on Robot and Human Interactive Communication. Kurashiki, Okayama Japan, September 2004. Jesse Hoey and James J. Little. Bayesian Clustering of Optical Flow Fields. In Proc. of International Conference on Computer Vision (ICCV), Nice, France, October 2003. Jesse Hoey. Clustering Contextual Facial Display Sequences. In Proc. of International Conference on Automatic Face and Gesture Recognition (FG), Washington, DC, May 2002. Pantelis Elinas, Jesse Hoey, Darrell Lahey, Jeff Montgomery, Don Murray, Stephen Se and James J. Little Waiting with Jose, a vision based mobile robot. In Proc. International Conference on Robotics and Automation (ICRA) Washington, DC, May 2002. Jesse Hoey. Hierarchical unsupervised learning of facial expression categories. In Proc. Workshop on detection and recognition of events in video, Vancouver, BC, July, 2001. Jesse Hoey and James J. Little. Representation and recognition of complex human motion. In Proc. of International Conference on Computer Vision and Pattern Recognition (CVPR), Hilton Head, SC, June 2000. Robert St-Aubin, Jesse Hoey, and Craig Boutilier. APRICODD: Approximate policy construction using decision diagrams. In Proc. Neural Information Processing Systems (NIPS) 14, 2000. Jesse Hoey, Robert St-Aubin, Alan Hu, and Craig Boutilier. SPUDD: Stochastic planning using decision diagrams. In Proc. of Uncertainty in Artificial Intelligence (UAI), Stockholm, 1999. Invited Talks and Demonstrations: --------------------------------- Assistive Technology and POMDPs March 18th, 2005 Presentation to AI group, University of Waterloo, Waterloo, Ontario Learning Models of Human Behavior using a Value Directed Approach June 9th 2004 IRIS Machine Learning Workshop 2004, Ottawa, Ontario University of British Columbia downtown campus opening gala July 10th 2002 Demonstrated Jose, the robotic waiter, to senior UBC officials. Vicky Gabereau Show. Canadian Broadcasting Corporation (CBC). September 20th 2002 Live television robotic waiter demonstration. UBC awards night gala reception June 3rd 2002 Robotic waiter demonstration. Advanced Systems Institute of British Columbia Technology Exchange. March 23rd 2002 Robotic waiter demonstration. UBC Department of Computer Science Open House January 17th 2002 Robotic waiter demonstration. AAAI Hors D'oe uvres Anyone? Mobile Robot Competition August 4th-9th 2001 Winning entry in the robotic waiter competition. Three demonstrations. Skills: ------- Programming Experience: Designed and wrote several large projects in C/C++ in both Unix and Microsoft Windows environments. Included writing code to communicate with hardware devices, and computer graphics user interfaces with OpenGL. OpenGL work further involved using graphics hardware for fast image computation, such as image re-scaling and optical flow computation. Expert in Matlab programming for computation and data visualization. Extensive experience programming in Java, FORTRAN, Visual Basic, and Scheme. Working knowledge of Prolog. Internet design work includes HTML, Javascript, CGI and Perl. Hardware Experience: Worked extensively with single-lens and stereo cameras. Worked on IEEE 1394 (Firewire) and s-video interfaces. Experienced working with a RWI B14 mobile robot. Languages: ---------- Completely fluent in written and spoken English and French. Basic spoken Spanish. Interests: ---------- Childhood development, skiing, playing hockey, playing guitar, traveling, reading, brewing. References: ----------- Craig Boutilier Department of Computer Science, University of Toronto, +1 (416) 946-5714 cebly@cs.toronto.edu James J. Little Department of Computer Science, University of British Columbia, +1 (604) 822-4830 little@cs.ubc.ca Alex Mihailidis Department of Occupational Therapy, University of Toronto, +1 (416) 946-8565 alex.mihailidis@utoronto.ca David Lowe Department of Computer Science, University of British Columbia, +1 (604) 822-6254 lowe@cs.ubc.ca