Grant: Discovery Grants Program - Individual
University of British Columbia
$29,000.00
Analysis and interpretation of visual motion
FY 2007-08
Summary
The goal is to make machines see. Research consists of selecting a task and implementing proof-of-concept systems to study performance in both real and controlled experimental situations. The current objective is to understand how motion supports high-level interpretation tasks related to an object's identity, to its non-visual physical properties and, for an object that is an active agent, to its state and intentions. Most often, recognition is taken to be the 'sine qua non' of computer vision. In this context, motion becomes the tracking of high-level features of objects already recognized statically. The study of biological vision systems suggests that motion itself is the more primitive cue. But, what are the low-level features to track? How might the tracking of low-level features drive recognition (rather than follow it)? These are the basic scientific questions. We have developed specific tools and experimental environments to explore these questions. Our Active Measurement Facility (ACME) is an integrated robotic measurement facility. ACME is used to acquire images of objects in motion using controlled illumination and sensing. ACME, by design, supports careful, repeatable scientific experimentation. Recent work interprets the motion of active agents, both cooperative and competitive, in the context of hockey. A proof-of-concept system has been demonstrated. The system takes as input player motion trajectory data tracked from real game video and supported by knowledge of hockey strategy, game situation and specific player profiles. The system describes what happened in each identified situation, assesses the outcome, estimates when and where key play choices were made, and attempts to predict whether better alternatives were available to achieve understood "coaching" objectives. This is an important first step toward generalizable, motion-based "situation analysis."
Program
Discovery Grants Program - Individual
Recipient
Location
British Columbia, Canada
Recipient Type
Academia
Type
Grant
Source
NSERC's Awards Data
As published
Application id
3390-2006
Name nom
Woodham, Robert
Department département
Computer Science
Organization id
2
Institution établissement
University of British Columbia
Province en
British Columbia
Province fr
Colombie-Britannique
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2007
Competition year année de concours
2006
Award amount
29000
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
Discovery grants programs
Group fr
Programmes de subventions à la découverte
Committee code
331
Committee name en
Computing and Information Sciences - B
Committee name fr
Informatique - B
Area of application code
1207
Area of application group en
Advancement of knowledge
Area of application group fr
Avancement des connaissances
Area of application en
Information, computer and communication technologies
Area of application fr
Technologie de l'information, des ordinateurs et des réseaux de communications
Research subject code
2603
Research subject group en
Robotics
Research subject group fr
Robotique
Research subject en
Computer vision
Research subject fr
Vision artificielle
Partie
2
Nb partie
5
Application title
Analysis and interpretation of visual motion
Keyword
NA
Application summary
The goal is to make machines see. Research consists of selecting a task and implementing proof-of-concept systems to study performance in both real and controlled experimental situations. The current objective is to understand how motion supports high-level interpretation tasks related to an object's identity, to its non-visual physical properties and, for an object that is an active agent, to its state and intentions. Most often, recognition is taken to be the 'sine qua non' of computer vision. In this context, motion becomes the tracking of high-level features of objects already recognized statically. The study of biological vision systems suggests that motion itself is the more primitive cue. But, what are the low-level features to track? How might the tracking of low-level features drive recognition (rather than follow it)? These are the basic scientific questions. We have developed specific tools and experimental environments to explore these questions. Our Active Measurement Facility (ACME) is an integrated robotic measurement facility. ACME is used to acquire images of objects in motion using controlled illumination and sensing. ACME, by design, supports careful, repeatable scientific experimentation. Recent work interprets the motion of active agents, both cooperative and competitive, in the context of hockey. A proof-of-concept system has been demonstrated. The system takes as input player motion trajectory data tracked from real game video and supported by knowledge of hockey strategy, game situation and specific player profiles. The system describes what happened in each identified situation, assesses the outcome, estimates when and where key play choices were made, and attempts to predict whether better alternatives were available to achieve understood "coaching" objectives. This is an important first step toward generalizable, motion-based "situation analysis."