Attention and Anticipation in Fast Visual-Inertial Navigation
We study a visual-inertial navigation (VIN) problem in which a robot needs to estimate its state using an on-board camera and an inertial sensor, without any prior knowledge of the external environment. We consider the case in which the robot can allocate limited resources to VIN, due to tight compu...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | English |
| Published: |
Institute of Electrical and Electronics Engineers (IEEE)
2020
|
| Online Access: | https://hdl.handle.net/1721.1/126592 |
| _version_ | 1826198915976790016 |
|---|---|
| author | Carlone, Luca Karaman, Sertac |
| author2 | Massachusetts Institute of Technology. Department of Aeronautics and Astronautics |
| author_facet | Massachusetts Institute of Technology. Department of Aeronautics and Astronautics Carlone, Luca Karaman, Sertac |
| author_sort | Carlone, Luca |
| collection | MIT |
| description | We study a visual-inertial navigation (VIN) problem in which a robot needs to estimate its state using an on-board camera and an inertial sensor, without any prior knowledge of the external environment. We consider the case in which the robot can allocate limited resources to VIN, due to tight computational constraints. Therefore, we answer the following question: under limited resources, what are the most relevant visual cues to maximize the performance of VIN? Our approach has four key ingredients. First, it is task-driven, in that the selection of the visual cues is guided by a metric quantifying the VIN performance. Second, it exploits the notion of anticipation, since it uses a simplified model for forward-simulation of robot dynamics, predicting the utility of a set of visual cues over a future time horizon. Third, it is efficient and easy to implement, since it leads to a greedy algorithm for the selection of the most relevant visual cues. Fourth, it provides formal performance guarantees: we leverage submodularity to prove that the greedy selection cannot be far from the optimal (combinatorial) selection. Simulations and real experiments on agile drones show that our approach ensures state-of-The-Art VIN performance while maintaining a lean processing time. In the easy scenarios, our approach outperforms appearance-based feature selection in terms of localization errors. In the most challenging scenarios, it enables accurate VIN while appearance-based feature selection fails to track robot's motion during aggressive maneuvers. |
| first_indexed | 2024-09-23T11:11:56Z |
| format | Article |
| id | mit-1721.1/126592 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T11:11:56Z |
| publishDate | 2020 |
| publisher | Institute of Electrical and Electronics Engineers (IEEE) |
| record_format | dspace |
| spelling | mit-1721.1/1265922022-09-27T17:47:00Z Attention and Anticipation in Fast Visual-Inertial Navigation Carlone, Luca Karaman, Sertac Massachusetts Institute of Technology. Department of Aeronautics and Astronautics Massachusetts Institute of Technology. Laboratory for Information and Decision Systems We study a visual-inertial navigation (VIN) problem in which a robot needs to estimate its state using an on-board camera and an inertial sensor, without any prior knowledge of the external environment. We consider the case in which the robot can allocate limited resources to VIN, due to tight computational constraints. Therefore, we answer the following question: under limited resources, what are the most relevant visual cues to maximize the performance of VIN? Our approach has four key ingredients. First, it is task-driven, in that the selection of the visual cues is guided by a metric quantifying the VIN performance. Second, it exploits the notion of anticipation, since it uses a simplified model for forward-simulation of robot dynamics, predicting the utility of a set of visual cues over a future time horizon. Third, it is efficient and easy to implement, since it leads to a greedy algorithm for the selection of the most relevant visual cues. Fourth, it provides formal performance guarantees: we leverage submodularity to prove that the greedy selection cannot be far from the optimal (combinatorial) selection. Simulations and real experiments on agile drones show that our approach ensures state-of-The-Art VIN performance while maintaining a lean processing time. In the easy scenarios, our approach outperforms appearance-based feature selection in terms of localization errors. In the most challenging scenarios, it enables accurate VIN while appearance-based feature selection fails to track robot's motion during aggressive maneuvers. 2020-08-14T19:18:50Z 2020-08-14T19:18:50Z 2017-07 2019-10-29T15:16:49Z Article http://purl.org/eprint/type/JournalArticle https://hdl.handle.net/1721.1/126592 Carlone, Luca and Sertac Karaman. “Attention and Anticipation in Fast Visual-Inertial Navigation.” Paper presented at the 2017 IEEE International Conference on Robotics and Automation (ICRA), Singapore, 29 May-3 June 2017, IEEE © 2017 The Author(s) en 10.1109/TRO.2018.2872402 2017 IEEE International Conference on Robotics and Automation (ICRA) Creative Commons Attribution-Noncommercial-Share Alike http://creativecommons.org/licenses/by-nc-sa/4.0/ application/pdf Institute of Electrical and Electronics Engineers (IEEE) arXiv |
| spellingShingle | Carlone, Luca Karaman, Sertac Attention and Anticipation in Fast Visual-Inertial Navigation |
| title | Attention and Anticipation in Fast Visual-Inertial Navigation |
| title_full | Attention and Anticipation in Fast Visual-Inertial Navigation |
| title_fullStr | Attention and Anticipation in Fast Visual-Inertial Navigation |
| title_full_unstemmed | Attention and Anticipation in Fast Visual-Inertial Navigation |
| title_short | Attention and Anticipation in Fast Visual-Inertial Navigation |
| title_sort | attention and anticipation in fast visual inertial navigation |
| url | https://hdl.handle.net/1721.1/126592 |
| work_keys_str_mv | AT carloneluca attentionandanticipationinfastvisualinertialnavigation AT karamansertac attentionandanticipationinfastvisualinertialnavigation |