Curriculum-based humanoid robot identification using large-scale human motion database
Identifying an accurate dynamics model remains challenging for humanoid robots. The difficulty is mainly due to the following two points. First, a good initial model is required to evaluate the feasibility of motions for data acquisition. Second, a highly nonlinear optimization problem needs to be s...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2023-11-01
|
| Series: | Frontiers in Robotics and AI |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/frobt.2023.1282299/full |
| _version_ | 1797447751483523072 |
|---|---|
| author | Sunhwi Kang Koji Ishihara Norikazu Sugimoto Jun Morimoto Jun Morimoto |
| author_facet | Sunhwi Kang Koji Ishihara Norikazu Sugimoto Jun Morimoto Jun Morimoto |
| author_sort | Sunhwi Kang |
| collection | DOAJ |
| description | Identifying an accurate dynamics model remains challenging for humanoid robots. The difficulty is mainly due to the following two points. First, a good initial model is required to evaluate the feasibility of motions for data acquisition. Second, a highly nonlinear optimization problem needs to be solved to design movements to acquire the identification data. To cope with the first point, in this paper, we propose a curriculum of identification to gradually learn an accurate dynamics model from an unreliable initial model. For the second point, we propose using a large-scale human motion database to efficiently design the humanoid movements for the parameter identification. The contribution of our study is developing a humanoid identification method that does not require the good initial model and does not need to solve the highly nonlinear optimization problem. We showed that our curriculum-based approach was able to more efficiently identify humanoid model parameters than a method that just randomly picked reference motions for identification. We evaluated our proposed method in a simulation experiment and demonstrated that our curriculum was led to obtain a wide variety of motion data for efficient parameter estimation. Consequently, our approach successfully identified an accurate model of an 18-DoF, simulated upper-body humanoid robot. |
| first_indexed | 2024-03-09T14:01:29Z |
| format | Article |
| id | doaj.art-f062fd3c3c2545d693444dbe7565322d |
| institution | Directory Open Access Journal |
| issn | 2296-9144 |
| language | English |
| last_indexed | 2024-03-09T14:01:29Z |
| publishDate | 2023-11-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Robotics and AI |
| spelling | doaj.art-f062fd3c3c2545d693444dbe7565322d2023-11-30T08:26:08ZengFrontiers Media S.A.Frontiers in Robotics and AI2296-91442023-11-011010.3389/frobt.2023.12822991282299Curriculum-based humanoid robot identification using large-scale human motion databaseSunhwi Kang0Koji Ishihara1Norikazu Sugimoto2Jun Morimoto3Jun Morimoto4Department of Brain Robot Interface, Computational Neuroscience Laboratories, Advanced Telecommunications Research Institute International (ATR), Kyoto, JapanDepartment of Brain Robot Interface, Computational Neuroscience Laboratories, Advanced Telecommunications Research Institute International (ATR), Kyoto, JapanDepartment of Brain Robot Interface, Computational Neuroscience Laboratories, Advanced Telecommunications Research Institute International (ATR), Kyoto, JapanDepartment of Brain Robot Interface, Computational Neuroscience Laboratories, Advanced Telecommunications Research Institute International (ATR), Kyoto, JapanGraduate School of Informatics, Kyoto University, Kyoto, JapanIdentifying an accurate dynamics model remains challenging for humanoid robots. The difficulty is mainly due to the following two points. First, a good initial model is required to evaluate the feasibility of motions for data acquisition. Second, a highly nonlinear optimization problem needs to be solved to design movements to acquire the identification data. To cope with the first point, in this paper, we propose a curriculum of identification to gradually learn an accurate dynamics model from an unreliable initial model. For the second point, we propose using a large-scale human motion database to efficiently design the humanoid movements for the parameter identification. The contribution of our study is developing a humanoid identification method that does not require the good initial model and does not need to solve the highly nonlinear optimization problem. We showed that our curriculum-based approach was able to more efficiently identify humanoid model parameters than a method that just randomly picked reference motions for identification. We evaluated our proposed method in a simulation experiment and demonstrated that our curriculum was led to obtain a wide variety of motion data for efficient parameter estimation. Consequently, our approach successfully identified an accurate model of an 18-DoF, simulated upper-body humanoid robot.https://www.frontiersin.org/articles/10.3389/frobt.2023.1282299/fullhuman motion databasehumanoid robotsmotion retargetingsystem identificationdynamics model |
| spellingShingle | Sunhwi Kang Koji Ishihara Norikazu Sugimoto Jun Morimoto Jun Morimoto Curriculum-based humanoid robot identification using large-scale human motion database Frontiers in Robotics and AI human motion database humanoid robots motion retargeting system identification dynamics model |
| title | Curriculum-based humanoid robot identification using large-scale human motion database |
| title_full | Curriculum-based humanoid robot identification using large-scale human motion database |
| title_fullStr | Curriculum-based humanoid robot identification using large-scale human motion database |
| title_full_unstemmed | Curriculum-based humanoid robot identification using large-scale human motion database |
| title_short | Curriculum-based humanoid robot identification using large-scale human motion database |
| title_sort | curriculum based humanoid robot identification using large scale human motion database |
| topic | human motion database humanoid robots motion retargeting system identification dynamics model |
| url | https://www.frontiersin.org/articles/10.3389/frobt.2023.1282299/full |
| work_keys_str_mv | AT sunhwikang curriculumbasedhumanoidrobotidentificationusinglargescalehumanmotiondatabase AT kojiishihara curriculumbasedhumanoidrobotidentificationusinglargescalehumanmotiondatabase AT norikazusugimoto curriculumbasedhumanoidrobotidentificationusinglargescalehumanmotiondatabase AT junmorimoto curriculumbasedhumanoidrobotidentificationusinglargescalehumanmotiondatabase AT junmorimoto curriculumbasedhumanoidrobotidentificationusinglargescalehumanmotiondatabase |