Learning Collective Crowd Behaviors with Dynamic Pedestrian-Agents
Collective behaviors characterize the intrinsic dynamics of the crowds. Automatically understanding collective crowd behaviors has important applications to video surveillance, traffic management and crowd control, while it is closely related to scientific fields such as statistical physics and biol...
| Main Authors: | , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | English |
| Published: |
Springer US
2016
|
| Online Access: | http://hdl.handle.net/1721.1/103360 https://orcid.org/0000-0002-3570-4396 |
| _version_ | 1826210993165828096 |
|---|---|
| author | Zhou, Bolei Tang, Xiaoou Wang, Xiaogang |
| author2 | Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science |
| author_facet | Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science Zhou, Bolei Tang, Xiaoou Wang, Xiaogang |
| author_sort | Zhou, Bolei |
| collection | MIT |
| description | Collective behaviors characterize the intrinsic dynamics of the crowds. Automatically understanding collective crowd behaviors has important applications to video surveillance, traffic management and crowd control, while it is closely related to scientific fields such as statistical physics and biology. In this paper, a new mixture model of dynamic pedestrian-Agents (MDA) is proposed to learn the collective behavior patterns of pedestrians in crowded scenes from video sequences. From agent-based modeling, each pedestrian in the crowd is driven by a dynamic pedestrian-agent, which is a linear dynamic system with initial and termination states reflecting the pedestrian’s belief of the starting point and the destination. The whole crowd is then modeled as a mixture of dynamic pedestrian-agents. Once the model parameters are learned from the trajectories extracted from videos, MDA can simulate the crowd behaviors. It can also infer the past behaviors and predict the future behaviors of pedestrians given their partially observed trajectories, and classify them different pedestrian behaviors. The effectiveness of MDA and its applications are demonstrated by qualitative and quantitative experiments on various video surveillance sequences. |
| first_indexed | 2024-09-23T14:59:01Z |
| format | Article |
| id | mit-1721.1/103360 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T14:59:01Z |
| publishDate | 2016 |
| publisher | Springer US |
| record_format | dspace |
| spelling | mit-1721.1/1033602022-10-01T23:45:29Z Learning Collective Crowd Behaviors with Dynamic Pedestrian-Agents Zhou, Bolei Tang, Xiaoou Wang, Xiaogang Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science Zhou, Bolei Collective behaviors characterize the intrinsic dynamics of the crowds. Automatically understanding collective crowd behaviors has important applications to video surveillance, traffic management and crowd control, while it is closely related to scientific fields such as statistical physics and biology. In this paper, a new mixture model of dynamic pedestrian-Agents (MDA) is proposed to learn the collective behavior patterns of pedestrians in crowded scenes from video sequences. From agent-based modeling, each pedestrian in the crowd is driven by a dynamic pedestrian-agent, which is a linear dynamic system with initial and termination states reflecting the pedestrian’s belief of the starting point and the destination. The whole crowd is then modeled as a mixture of dynamic pedestrian-agents. Once the model parameters are learned from the trajectories extracted from videos, MDA can simulate the crowd behaviors. It can also infer the past behaviors and predict the future behaviors of pedestrians given their partially observed trajectories, and classify them different pedestrian behaviors. The effectiveness of MDA and its applications are demonstrated by qualitative and quantitative experiments on various video surveillance sequences. Research Grants Council (Hong Kong, China) (Project No. CUHK417110) Research Grants Council (Hong Kong, China) (Project No. CUHK417011) Research Grants Council (Hong Kong, China) (Project No. CUHK 429412). 2016-06-27T19:30:55Z 2016-06-27T19:30:55Z 2014-06 2013-09 2016-05-23T12:14:39Z Article http://purl.org/eprint/type/JournalArticle 0920-5691 1573-1405 http://hdl.handle.net/1721.1/103360 Zhou, Bolei, Xiaoou Tang, and Xiaogang Wang. "Learning Collective Crowd Behaviors with Dynamic Pedestrian-Agents." International Journal of Computer Vision 111:1 (January 2015), pp 50-68. https://orcid.org/0000-0002-3570-4396 en http://dx.doi.org/10.1007/s11263-014-0735-3 International Journal of Computer Vision Article is made available in accordance with the publisher's policy and may be subject to US copyright law. Please refer to the publisher's site for terms of use. Springer Science+Business Media New York application/pdf Springer US Springer US |
| spellingShingle | Zhou, Bolei Tang, Xiaoou Wang, Xiaogang Learning Collective Crowd Behaviors with Dynamic Pedestrian-Agents |
| title | Learning Collective Crowd Behaviors with Dynamic Pedestrian-Agents |
| title_full | Learning Collective Crowd Behaviors with Dynamic Pedestrian-Agents |
| title_fullStr | Learning Collective Crowd Behaviors with Dynamic Pedestrian-Agents |
| title_full_unstemmed | Learning Collective Crowd Behaviors with Dynamic Pedestrian-Agents |
| title_short | Learning Collective Crowd Behaviors with Dynamic Pedestrian-Agents |
| title_sort | learning collective crowd behaviors with dynamic pedestrian agents |
| url | http://hdl.handle.net/1721.1/103360 https://orcid.org/0000-0002-3570-4396 |
| work_keys_str_mv | AT zhoubolei learningcollectivecrowdbehaviorswithdynamicpedestrianagents AT tangxiaoou learningcollectivecrowdbehaviorswithdynamicpedestrianagents AT wangxiaogang learningcollectivecrowdbehaviorswithdynamicpedestrianagents |