A Framework for Recognizing Multi-Agent Action from Visual Evidence

A probabilistic framework forrepresenting and visually recognizing complex multi-agent action is presented. Motivated by work in model-based object recognition and designed for the recognition of action from visual evidence, the representation has three components: (1) temporal structure descriptions representing the temporal relationships between agent goals, (2) belief networks for probabilistically representing and recognizing individual agentgoals from visualevidence, and (3) belief networks automatically generated from the temporalstructure descriptions that supportthe recognition of the complex action. We describe our current work on recognizing American football plays from noisy trajectory data. 1

Authors

• Stephen S. Intille
• Aaron F. Bobick
• MIT Media Laboratory

Keywords

No keywords are indexed for this paper.