| Presentation | 2006-07-14 Recognition of human movements by using the ground reaction force and motion capture data Yuka ARIKI, Jun MORIMOTO, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | Motion capture data have been commonly used to generate human-like behaviours for humanoid robots and virtual characters. One of the key issues in behaviour generation studies is how to create a wide variety of motions by combining movement primitives, while each primitive is extracted from human. Previous studies used captured joint angles data for behaviour discrimination and for generating combined behaviours, but have not explicitly considered its physical plausibility. In this study, we consider using not only the motion capture data but also the ground reaction force information to generate physically plausible motions. We selected walking and running motions as the primitive behaviours in our study. We generate a combined behaviour from the two primitive motions, captured from human subjects. We selected frames for transition from one primitive to the other by finding the minimum distance between two frames of the joint angles and ground reaction force. As a result, we could generate a physically plausible motion. We also applied our proposed approach to discriminate the two primitive behaviours from a combined motion. We successfully discriminate two behaviours with using switching state-space models. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Motion recognition / Motion generation / Switching state-sapce models / motion capture |
| Paper # | NC2006-44 |
| Date of Issue |
| Conference Information | |
| Committee | NC |
|---|---|
| Conference Date | 2006/7/7(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 | Neurocomputing (NC) |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Recognition of human movements by using the ground reaction force and motion capture data |
| Sub Title (in English) | |
| Keyword(1) | Motion recognition |
| Keyword(2) | Motion generation |
| Keyword(3) | Switching state-sapce models |
| Keyword(4) | motion capture |
| 1st Author's Name | Yuka ARIKI |
| 1st Author's Affiliation | Nara Institute of Science and Technology:ATR, CNS() |
| 2nd Author's Name | Jun MORIMOTO |
| 2nd Author's Affiliation | ATR, CNS:ICORP, JST |
| Date | 2006-07-14 |
| Paper # | NC2006-44 |
| Volume (vol) | vol.106 |
| Number (no) | 163 |
| Page | pp.pp.- |
| #Pages | 5 |
| Date of Issue | |