Propagation networks for model-based control under partial observation
There has been an increasing interest in learning dynamics simulators for model-based control. Compared with off-the-shelf physics engines, a learnable simulator can quickly adapt to unseen objects, scenes, and tasks. However, existing models like interaction networks only work for fully observable...
| Main Authors: | , , , , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2020
|
| Online Access: | https://hdl.handle.net/1721.1/126583 |
| _version_ | 1826190523914780672 |
|---|---|
| author | Li, Yunzhu Wu, Jiajun Zhu, Junyan Tenenbaum, Joshua B Torralba, Antonio Tedrake, Russell L |
| author2 | Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory |
| author_facet | Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory Li, Yunzhu Wu, Jiajun Zhu, Junyan Tenenbaum, Joshua B Torralba, Antonio Tedrake, Russell L |
| author_sort | Li, Yunzhu |
| collection | MIT |
| description | There has been an increasing interest in learning dynamics simulators for model-based control. Compared with off-the-shelf physics engines, a learnable simulator can quickly adapt to unseen objects, scenes, and tasks. However, existing models like interaction networks only work for fully observable systems; they also only consider pairwise interactions within a single time step, both restricting their use in practical systems. We introduce Propagation Networks (PropNet), a differentiable, learnable dynamics model that handles partially observable scenarios and enables instantaneous propagation of signals beyond pairwise interactions. With these innovations, our propagation networks not only outperform current learnable physics engines in forward simulation, but also achieves superior performance on various control tasks. Compared with existing deep reinforcement learning algorithms, model-based control with propagation networks is more accurate, efficient, and generalizable to novel, partially observable scenes and tasks. |
| first_indexed | 2024-09-23T08:41:35Z |
| format | Article |
| id | mit-1721.1/126583 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T08:41:35Z |
| publishDate | 2020 |
| publisher | IEEE |
| record_format | dspace |
| spelling | mit-1721.1/1265832022-09-23T13:53:36Z Propagation networks for model-based control under partial observation Li, Yunzhu Wu, Jiajun Zhu, Junyan Tenenbaum, Joshua B Torralba, Antonio Tedrake, Russell L Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory Massachusetts Institute of Technology. Department of Brain and Cognitive Sciences Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science There has been an increasing interest in learning dynamics simulators for model-based control. Compared with off-the-shelf physics engines, a learnable simulator can quickly adapt to unseen objects, scenes, and tasks. However, existing models like interaction networks only work for fully observable systems; they also only consider pairwise interactions within a single time step, both restricting their use in practical systems. We introduce Propagation Networks (PropNet), a differentiable, learnable dynamics model that handles partially observable scenarios and enables instantaneous propagation of signals beyond pairwise interactions. With these innovations, our propagation networks not only outperform current learnable physics engines in forward simulation, but also achieves superior performance on various control tasks. Compared with existing deep reinforcement learning algorithms, model-based control with propagation networks is more accurate, efficient, and generalizable to novel, partially observable scenes and tasks. Charles Stark Draper Laboratory. Sponsor Award (SC001-0000001002) United States. National Aeronautics and Space Administration Sponsor Award (NNX16AC49A) National Science Foundation (U.S.) (Grant 1524817) United States. Defense Advanced Research Projects Agency. Explainable Artificial Intelligence (Grant FA8750-18-C000) United States. Office of Naval Research. Multidisciplinary University Research Initiative (Grant N00014-16-1-2007) 2020-08-14T13:23:08Z 2020-08-14T13:23:08Z 2019-05 2019-10-09T12:04:27Z Article http://purl.org/eprint/type/ConferencePaper https://hdl.handle.net/1721.1/126583 Li, Yunzhu et al. “Propagation networks for model-based control under partial observation.” Paper presented at the 2019 International Conference on Robotics and Automation (ICRA), Montréal, Québec, May 20-24, 2019, IEEE © 2019 The Author(s) en 10.1109/ICRA.2019.8793509 2019 International Conference on Robotics and Automation (ICRA) Creative Commons Attribution-Noncommercial-Share Alike http://creativecommons.org/licenses/by-nc-sa/4.0/ application/pdf IEEE arXiv |
| spellingShingle | Li, Yunzhu Wu, Jiajun Zhu, Junyan Tenenbaum, Joshua B Torralba, Antonio Tedrake, Russell L Propagation networks for model-based control under partial observation |
| title | Propagation networks for model-based control under partial observation |
| title_full | Propagation networks for model-based control under partial observation |
| title_fullStr | Propagation networks for model-based control under partial observation |
| title_full_unstemmed | Propagation networks for model-based control under partial observation |
| title_short | Propagation networks for model-based control under partial observation |
| title_sort | propagation networks for model based control under partial observation |
| url | https://hdl.handle.net/1721.1/126583 |
| work_keys_str_mv | AT liyunzhu propagationnetworksformodelbasedcontrolunderpartialobservation AT wujiajun propagationnetworksformodelbasedcontrolunderpartialobservation AT zhujunyan propagationnetworksformodelbasedcontrolunderpartialobservation AT tenenbaumjoshuab propagationnetworksformodelbasedcontrolunderpartialobservation AT torralbaantonio propagationnetworksformodelbasedcontrolunderpartialobservation AT tedrakerusselll propagationnetworksformodelbasedcontrolunderpartialobservation |