Paper Abstract and Keywords |
Presentation |
2016-12-08 06:55
Sign Language Recognition Performance Improvements by Feature Elements of Hidden Markov Model Hirotoshi Shibata, Tatsunori Ozawa, Hiromitsu Nishimura, Hiroshi Tanaka (Kanagawa Institute of Technology), Daisuke Kobayashi, Michio Iwamoto, Syuji Kato (KCC corp.) |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
The authors have been investigating method of sign language recognition using colored gloves and optical camera. We are now applying HMM which has been reported high recognition performance for sign language recognition. So far, we have evaluated HMM which was learned shape of hand motion. However, recognition performance was insufficient. This paper reports the recognition results by considering the velocity of motion of hand, position of hand, and visible status of each finger. In addition, it is shown the recognition enhancement can be obtained by considering the reliability obtained by HMM which was learned each feature vector. |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
Sign Language Recognition / Colored Gloves / Optical Camera / Hidden Markov Model / Feature Elements / / / |
Reference Info. |
IEICE Tech. Rep. |
Paper # |
|
Date of Issue |
|
ISSN |
|
Download PDF |
|