Presentation | 2000/6/15 SP2000-13 Speaker Independent Phoneme Recognition Based on A Speaker Cluster Estimated From Input Speech Motoyuki Suzuki, Shozo Makino, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | A kind of speaker adaptation step is carried out in many speech recognition systems. How-ever, in recognition system where amount of speech data to be recognized is small or where a speaker is frequently changed, speaker-independent speech recognition system is essential. In this paper, we have proposed speaker-independent speech recognition system with high performance using HMnet made by SSS-free. In this system, speaker cluster is made using distance between speakers, and speech recog-nition is carried out using an acoustic model constructed from speaker's model belonging to a speaker cluster. The proposed system showed about 4% higher recognition accuracy than speaker-dependent one with traditional speaker adaptation method. We have also proposed automatic adjustment method of threshold used in the making step of speaker chuster. The performance using new adjustment method is as same as the maximum performance using constant threshold. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Speaker-independent speech recognition / Distance between speakers / SSS-free / HMnet |
Paper # | SP2000-13 |
Date of Issue |
Conference Information | |
Committee | SP |
---|---|
Conference Date | 2000/6/15(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 | Speech (SP) |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | SP2000-13 Speaker Independent Phoneme Recognition Based on A Speaker Cluster Estimated From Input Speech |
Sub Title (in English) | |
Keyword(1) | Speaker-independent speech recognition |
Keyword(2) | Distance between speakers |
Keyword(3) | SSS-free |
Keyword(4) | HMnet |
1st Author's Name | Motoyuki Suzuki |
1st Author's Affiliation | Computer Center. Tohoku University() |
2nd Author's Name | Shozo Makino |
2nd Author's Affiliation | Computer Center. Tohoku University |
Date | 2000/6/15 |
Paper # | SP2000-13 |
Volume (vol) | vol.100 |
Number (no) | 136 |
Page | pp.pp.- |
#Pages | 8 |
Date of Issue |