Presentation | 2005-10-28 Learning, Recognition and Generation of Time-series Patterns Based on Self-organizing Segmentation Shogo OKADA, Osamu HASEGAWA, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | This study is intended to realize a flexible learning mechanism that can be possible to unsupervised learning, semi-supervised learning, incremental learning, recognize and generate time-series patterns (dynamic patterns) in the real world. In addition, this mechanism can learn new patterns incrementally. The mechanism divides self-organized patterns into primitives by mixture-of-experts (MoE) system. Each expert learns the pattern primitives. Experts of the MoE are small non-monotonous neural networks. Learning patterns are expressed as a permutation of primitives that are output by the MoE, recognized, and then generated by applying the permutation and DPmatching. We confirmed the effectiveness of this mechanism by two experiments that used gestures directly from the motion without any structural information. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | pattern recognition / gesture recognition / self-organization / neural network |
Paper # | PRMU2005-103 |
Date of Issue |
Conference Information | |
Committee | PRMU |
---|---|
Conference Date | 2005/10/21(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) | Learning, Recognition and Generation of Time-series Patterns Based on Self-organizing Segmentation |
Sub Title (in English) | |
Keyword(1) | pattern recognition |
Keyword(2) | gesture recognition |
Keyword(3) | self-organization |
Keyword(4) | neural network |
1st Author's Name | Shogo OKADA |
1st Author's Affiliation | Tokyo Institute of Technology() |
2nd Author's Name | Osamu HASEGAWA |
2nd Author's Affiliation | Tokyo Institute of Technology:PRESTO, JST |
Date | 2005-10-28 |
Paper # | PRMU2005-103 |
Volume (vol) | vol.105 |
Number (no) | 375 |
Page | pp.pp.- |
#Pages | 6 |
Date of Issue |