DualSMC: Tunneling Differentiable Filtering and Planning under Continuous POMDPs
© 2020 Inst. Sci. inf., Univ. Defence in Belgrade. All rights reserved. A major difficulty of solving continuous POMDPs is to infer the multi-modal distribution of the unobserved true states and to make the planning algorithm dependent on the perceived uncertainty. We cast POMDP filtering and planni...
| Main Authors: | , , , , , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | English |
| Published: |
International Joint Conferences on Artificial Intelligence Organization
2021
|
| Online Access: | https://hdl.handle.net/1721.1/138359 |
| _version_ | 1811084178326814720 |
|---|---|
| author | Wang, Yunbo Liu, Bo Wu, Jiajun Zhu, Yuke Du, Simon S Fei-Fei, Li Tenenbaum, Joshua B |
| author2 | Massachusetts Institute of Technology. Department of Brain and Cognitive Sciences |
| author_facet | Massachusetts Institute of Technology. Department of Brain and Cognitive Sciences Wang, Yunbo Liu, Bo Wu, Jiajun Zhu, Yuke Du, Simon S Fei-Fei, Li Tenenbaum, Joshua B |
| author_sort | Wang, Yunbo |
| collection | MIT |
| description | © 2020 Inst. Sci. inf., Univ. Defence in Belgrade. All rights reserved. A major difficulty of solving continuous POMDPs is to infer the multi-modal distribution of the unobserved true states and to make the planning algorithm dependent on the perceived uncertainty. We cast POMDP filtering and planning problems as two closely related Sequential Monte Carlo (SMC) processes, one over the real states and the other over the future optimal trajectories, and combine the merits of these two parts in a new model named the DualSMC network. In particular, we first introduce an adversarial particle filter that leverages the adversarial relationship between its internal components. Based on the filtering results, we then propose a planning algorithm that extends the previous SMC planning approach [Piche et al., 2018] to continuous POMDPs with an uncertainty-dependent policy. Crucially, not only can DualSMC handle complex observations such as image input but also it remains highly interpretable. It is shown to be effective in three continuous POMDP domains: the floor positioning domain, the 3D light-dark navigation domain, and a modified Reacher domain. |
| first_indexed | 2024-09-23T12:46:08Z |
| format | Article |
| id | mit-1721.1/138359 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T12:46:08Z |
| publishDate | 2021 |
| publisher | International Joint Conferences on Artificial Intelligence Organization |
| record_format | dspace |
| spelling | mit-1721.1/1383592023-02-03T20:00:47Z DualSMC: Tunneling Differentiable Filtering and Planning under Continuous POMDPs Wang, Yunbo Liu, Bo Wu, Jiajun Zhu, Yuke Du, Simon S Fei-Fei, Li Tenenbaum, Joshua B Massachusetts Institute of Technology. Department of Brain and Cognitive Sciences Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory © 2020 Inst. Sci. inf., Univ. Defence in Belgrade. All rights reserved. A major difficulty of solving continuous POMDPs is to infer the multi-modal distribution of the unobserved true states and to make the planning algorithm dependent on the perceived uncertainty. We cast POMDP filtering and planning problems as two closely related Sequential Monte Carlo (SMC) processes, one over the real states and the other over the future optimal trajectories, and combine the merits of these two parts in a new model named the DualSMC network. In particular, we first introduce an adversarial particle filter that leverages the adversarial relationship between its internal components. Based on the filtering results, we then propose a planning algorithm that extends the previous SMC planning approach [Piche et al., 2018] to continuous POMDPs with an uncertainty-dependent policy. Crucially, not only can DualSMC handle complex observations such as image input but also it remains highly interpretable. It is shown to be effective in three continuous POMDP domains: the floor positioning domain, the 3D light-dark navigation domain, and a modified Reacher domain. 2021-12-07T19:14:34Z 2021-12-07T19:14:34Z 2020 2021-12-07T19:08:50Z Article http://purl.org/eprint/type/JournalArticle https://hdl.handle.net/1721.1/138359 Wang, Yunbo, Liu, Bo, Wu, Jiajun, Zhu, Yuke, Du, Simon S et al. 2020. "DualSMC: Tunneling Differentiable Filtering and Planning under Continuous POMDPs." IJCAI International Joint Conference on Artificial Intelligence, 2021-January. en 10.24963/IJCAI.2020/579 IJCAI International Joint Conference on Artificial Intelligence Creative Commons Attribution-Noncommercial-Share Alike http://creativecommons.org/licenses/by-nc-sa/4.0/ application/pdf International Joint Conferences on Artificial Intelligence Organization arXiv |
| spellingShingle | Wang, Yunbo Liu, Bo Wu, Jiajun Zhu, Yuke Du, Simon S Fei-Fei, Li Tenenbaum, Joshua B DualSMC: Tunneling Differentiable Filtering and Planning under Continuous POMDPs |
| title | DualSMC: Tunneling Differentiable Filtering and Planning under Continuous POMDPs |
| title_full | DualSMC: Tunneling Differentiable Filtering and Planning under Continuous POMDPs |
| title_fullStr | DualSMC: Tunneling Differentiable Filtering and Planning under Continuous POMDPs |
| title_full_unstemmed | DualSMC: Tunneling Differentiable Filtering and Planning under Continuous POMDPs |
| title_short | DualSMC: Tunneling Differentiable Filtering and Planning under Continuous POMDPs |
| title_sort | dualsmc tunneling differentiable filtering and planning under continuous pomdps |
| url | https://hdl.handle.net/1721.1/138359 |
| work_keys_str_mv | AT wangyunbo dualsmctunnelingdifferentiablefilteringandplanningundercontinuouspomdps AT liubo dualsmctunnelingdifferentiablefilteringandplanningundercontinuouspomdps AT wujiajun dualsmctunnelingdifferentiablefilteringandplanningundercontinuouspomdps AT zhuyuke dualsmctunnelingdifferentiablefilteringandplanningundercontinuouspomdps AT dusimons dualsmctunnelingdifferentiablefilteringandplanningundercontinuouspomdps AT feifeili dualsmctunnelingdifferentiablefilteringandplanningundercontinuouspomdps AT tenenbaumjoshuab dualsmctunnelingdifferentiablefilteringandplanningundercontinuouspomdps |