Modeling and estimating persistent motion with geometric flows
We propose a principled framework to model persistent motion in dynamic scenes. In contrast to previous efforts on object tracking and optical flow estimation that focus on local motion, we primarily aim at inferring a global model of persistent and collective dynamics. With this in mind, we first i...
| Main Authors: | , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | en_US |
| Published: |
Institute of Electrical and Electronics Engineers (IEEE)
2012
|
| Online Access: | http://hdl.handle.net/1721.1/71897 https://orcid.org/0000-0003-4844-3495 https://orcid.org/0000-0002-6192-2207 |
| _version_ | 1826203908084596736 |
|---|---|
| author | Lin, Dahua Grimson, Eric Fisher, John W., III |
| author2 | Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory |
| author_facet | Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory Lin, Dahua Grimson, Eric Fisher, John W., III |
| author_sort | Lin, Dahua |
| collection | MIT |
| description | We propose a principled framework to model persistent motion in dynamic scenes. In contrast to previous efforts on object tracking and optical flow estimation that focus on local motion, we primarily aim at inferring a global model of persistent and collective dynamics. With this in mind, we first introduce the concept of geometric flow that describes motion simultaneously over space and time, and derive a vector space representation based on Lie algebra. We then extend it to model complex motion by combining multiple flows in a geometrically consistent manner. Taking advantage of the linear nature of this representation, we formulate a stochastic flow model, and incorporate a Gaussian process to capture the spatial coherence more effectively. This model leads to an efficient and robust algorithm that can integrate both point pairs and frame differences in motion estimation. We conducted experiments on different types of videos. The results clearly demonstrate that the proposed approach is effective in modeling persistent motion. |
| first_indexed | 2024-09-23T12:45:15Z |
| format | Article |
| id | mit-1721.1/71897 |
| institution | Massachusetts Institute of Technology |
| language | en_US |
| last_indexed | 2024-09-23T12:45:15Z |
| publishDate | 2012 |
| publisher | Institute of Electrical and Electronics Engineers (IEEE) |
| record_format | dspace |
| spelling | mit-1721.1/718972022-09-28T09:52:32Z Modeling and estimating persistent motion with geometric flows Lin, Dahua Grimson, Eric Fisher, John W., III Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science Grimson, William E. Lin, Dahua Grimson, Eric Fisher, John W., III We propose a principled framework to model persistent motion in dynamic scenes. In contrast to previous efforts on object tracking and optical flow estimation that focus on local motion, we primarily aim at inferring a global model of persistent and collective dynamics. With this in mind, we first introduce the concept of geometric flow that describes motion simultaneously over space and time, and derive a vector space representation based on Lie algebra. We then extend it to model complex motion by combining multiple flows in a geometrically consistent manner. Taking advantage of the linear nature of this representation, we formulate a stochastic flow model, and incorporate a Gaussian process to capture the spatial coherence more effectively. This model leads to an efficient and robust algorithm that can integrate both point pairs and frame differences in motion estimation. We conducted experiments on different types of videos. The results clearly demonstrate that the proposed approach is effective in modeling persistent motion. United States. Army Research Office. Multidisciplinary University Research Initiative. (Heterogeneous Sensor Networks) (Award number W911NF-06-1-0076). United States. Air Force Office of Scientific Research (Award Number FA9550-06-1-0324) 2012-07-30T19:35:39Z 2012-07-30T19:35:39Z 2010-08 2010-06 Article http://purl.org/eprint/type/ConferencePaper 978-1-4244-6984-0 1063-6919 http://hdl.handle.net/1721.1/71897 Lin, Dahua, Eric Grimson, and John Fisher. “Modeling and Estimating Persistent Motion with Geometric Flows.” IEEE, 2010. 1–8. © Copyright 2010 IEEE https://orcid.org/0000-0003-4844-3495 https://orcid.org/0000-0002-6192-2207 en_US http://dx.doi.org/ 10.1109/CVPR.2010.5539848 2010 IEEE Conference on Computer Vision and Pattern Recognition 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. application/pdf Institute of Electrical and Electronics Engineers (IEEE) IEEE |
| spellingShingle | Lin, Dahua Grimson, Eric Fisher, John W., III Modeling and estimating persistent motion with geometric flows |
| title | Modeling and estimating persistent motion with geometric flows |
| title_full | Modeling and estimating persistent motion with geometric flows |
| title_fullStr | Modeling and estimating persistent motion with geometric flows |
| title_full_unstemmed | Modeling and estimating persistent motion with geometric flows |
| title_short | Modeling and estimating persistent motion with geometric flows |
| title_sort | modeling and estimating persistent motion with geometric flows |
| url | http://hdl.handle.net/1721.1/71897 https://orcid.org/0000-0003-4844-3495 https://orcid.org/0000-0002-6192-2207 |
| work_keys_str_mv | AT lindahua modelingandestimatingpersistentmotionwithgeometricflows AT grimsoneric modelingandestimatingpersistentmotionwithgeometricflows AT fisherjohnwiii modelingandestimatingpersistentmotionwithgeometricflows |