A Framework for Recognizing Multi-Agent Action from Visual Evidence
Abstract
A probabilistic framework for representing 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 agent goals from visual evidence, and (3) belief networks automatically generated from the temporal structure descriptions that support the recognition of the complex action. We describe our current work on recognizing American football plays from noisy trajectory data. 1 Keywords: action recognition, plan recognition, representing visual uncertainty 1 Introduction Evaluating whether an observed set of visual phenomena constitute a particular dynamic event requires representation and recognition of temporal relationships and uncertain information...
Cite
Text
Intille and Bobick. "A Framework for Recognizing Multi-Agent Action from Visual Evidence." AAAI Conference on Artificial Intelligence, 1999.Markdown
[Intille and Bobick. "A Framework for Recognizing Multi-Agent Action from Visual Evidence." AAAI Conference on Artificial Intelligence, 1999.](https://mlanthology.org/aaai/1999/intille1999aaai-framework/)BibTeX
@inproceedings{intille1999aaai-framework,
title = {{A Framework for Recognizing Multi-Agent Action from Visual Evidence}},
author = {Intille, Stephen S. and Bobick, Aaron F.},
booktitle = {AAAI Conference on Artificial Intelligence},
year = {1999},
pages = {518-525},
url = {https://mlanthology.org/aaai/1999/intille1999aaai-framework/}
}