| Presentation | 2005-09-22 Motion learning and generation of humanoids based on symbolization of gesture and voice instruction by HMM Kazuyuki SAKAMOTO, Tetsunari INAMURA, Naoki KOJO, Tomoyuki SONODA, Kei OKADA, Masayuki INABA, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | When humanoids observe and acquire the humans' daily-life behavior, objective imitation could not be achieved by simple copy of joint angles. It is necessary for achievement of aim of daily-life behaviors to consider focus points of the behaviors and conditions of constraint, and integrate them with motion patterns such as time-seires data of joint angles. Recently, motion capture systems are widely used as a method of teaching whole body motion, but there were no effective methods for teaching objective of the motions, learning of the motions and realization of behavior generation by real huamanoid robots. In this paper, we introduce an approach that conditions for achievement of behaviors' objective are described as focus points and conditions of constraint. And we propose an approach not only motion patterns but also focus points and conditions of constraint are taken into the learning, recognition and generation processes. We propose a teaching system which enables humanoid robots to integrate learning, recognition and generation processes based on combination of symbolizing technology of time-series patterns based on hidden Markov models and real-time and on-sight interface technology for teaching motion patterns and objectives of behavior. Finally, we show effectiveness of the proposed method through experiments by a real humanoid robot. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | focus point / condition of constraint / symbolization / teaching motion |
| Paper # | NLC2005-39,PRMU2005-66 |
| Date of Issue |
| Conference Information | |
| Committee | PRMU |
|---|---|
| Conference Date | 2005/9/15(1days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | |
| Chair | |
| Vice Chair | |
| Secretary | |
| Assistant | |
| Paper Information | |
| Registration To | Pattern Recognition and Media Understanding (PRMU) |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Motion learning and generation of humanoids based on symbolization of gesture and voice instruction by HMM |
| Sub Title (in English) | |
| Keyword(1) | focus point |
| Keyword(2) | condition of constraint |
| Keyword(3) | symbolization |
| Keyword(4) | teaching motion |
| 1st Author's Name | Kazuyuki SAKAMOTO |
| 1st Author's Affiliation | Graduate School of Information Science and Technology, the University of Tokyo() |
| 2nd Author's Name | Tetsunari INAMURA |
| 2nd Author's Affiliation | Graduate School of Information Science and Technology, the University of Tokyo |
| 3rd Author's Name | Naoki KOJO |
| 3rd Author's Affiliation | Graduate School of Information Science and Technology, the University of Tokyo |
| 4th Author's Name | Tomoyuki SONODA |
| 4th Author's Affiliation | Graduate School of Information Science and Technology, the University of Tokyo |
| 5th Author's Name | Kei OKADA |
| 5th Author's Affiliation | Graduate School of Information Science and Technology, the University of Tokyo |
| 6th Author's Name | Masayuki INABA |
| 6th Author's Affiliation | Graduate School of Information Science and Technology, the University of Tokyo |
| Date | 2005-09-22 |
| Paper # | NLC2005-39,PRMU2005-66 |
| Volume (vol) | vol.105 |
| Number (no) | 302 |
| Page | pp.pp.- |
| #Pages | 6 |
| Date of Issue | |