| Presentation | 2010-03-12 MSD-HMM for Occluded Hands in Sign Language Operation Shinji SAKO, Tadashi KITAMURA, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | In sign language recognition system, there are two main methods for capturing hands movement from signers. The first one is generally called the motion capture method which uses special sensors (e.g. CyberGlove[○!R]) capture the movement of the signer. The second one, which is called image-based approach relies on videos of signers to capture signs. In this paper, We adopt the second approach. While image-based approach has some advantages against sensor-based approach, it is difficult to overcome occlusion problems. Hidden Markov Model (HMM) have been widely used in sign language recognition. Also, in the field on speech recognition and speech synthesis, HMM have been used for acoustic modeling. In speech synthesis domain, Multi-space probability HMM (MSD-HMM) have been proposed as an extension of HMM that includes discrete and continuous HMM. We adopt MSD-HMM technique in order to overcome occlusion problems. This paper describes to use multi-space probability distribution (MSD) to model occluded hand positions jointly with visible hand positions. Experimental result obtained in Japanese sign language recognition indicate MSD provided almost equal to word recognition accuracy across conventional HMM method using interpolated trajectory of hand positions. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Hidden Markov Model / Sign Language Recognition / Multi-space distribution / Occlusion |
| Paper # | WIT2009-88 |
| Date of Issue |
| Conference Information | |
| Committee | WIT |
|---|---|
| Conference Date | 2010/3/5(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 | Well-being Information Technology(WIT) |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | MSD-HMM for Occluded Hands in Sign Language Operation |
| Sub Title (in English) | |
| Keyword(1) | Hidden Markov Model |
| Keyword(2) | Sign Language Recognition |
| Keyword(3) | Multi-space distribution |
| Keyword(4) | Occlusion |
| 1st Author's Name | Shinji SAKO |
| 1st Author's Affiliation | Department of Computer Science and Engineering Graduate School of Engineering, Nagoya Institute of Technology() |
| 2nd Author's Name | Tadashi KITAMURA |
| 2nd Author's Affiliation | Department of Computer Science and Engineering Graduate School of Engineering, Nagoya Institute of Technology |
| Date | 2010-03-12 |
| Paper # | WIT2009-88 |
| Volume (vol) | vol.109 |
| Number (no) | 467 |
| Page | pp.pp.- |
| #Pages | 6 |
| Date of Issue | |