Presentation | 1997/12/11 Unsupervised Speaker Normalization by Canonical Correlation Analysis to Phoneme Representation Vectors Miharu Sakuragi, Yasuo Ariki, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | Conventional speaker-independent HMMs ignore the speaker differences and collect speech data in an observation space. This causes a problem that the output probability distribution of the HMMs becomes vague so that, it deteriorates the recognition accuracy. To solve this problem, we construct the speaker subspace for an individual speaker and correlate them by o-space canonical correlation analysis between the standard speaker and input speaker. In order to remove the constraint that input speakers have to speak the same sentences as the standard speaker in the supervised normalization, we propose in this paper an unsupervised speaker normalization method which automatically segments the speech data into phoneme data by Viterbi decoding algorithm and then associates the mean feature vectors of phoneme data between speakers by o-space canonical correlation analysis. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | speaker-independent HMMs / supervised speaker normalization / unsupervised speaker normalization / Viterbe decoding algorithm |
Paper # | SP97-72 |
Date of Issue |
Conference Information | |
Committee | SP |
---|---|
Conference Date | 1997/12/11(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) | Unsupervised Speaker Normalization by Canonical Correlation Analysis to Phoneme Representation Vectors |
Sub Title (in English) | |
Keyword(1) | speaker-independent HMMs |
Keyword(2) | supervised speaker normalization |
Keyword(3) | unsupervised speaker normalization |
Keyword(4) | Viterbe decoding algorithm |
1st Author's Name | Miharu Sakuragi |
1st Author's Affiliation | Faculty of Science and Technology, Ryukoku University() |
2nd Author's Name | Yasuo Ariki |
2nd Author's Affiliation | Faculty of Science and Technology, Ryukoku University |
Date | 1997/12/11 |
Paper # | SP97-72 |
Volume (vol) | vol.97 |
Number (no) | 441 |
Page | pp.pp.- |
#Pages | 6 |
Date of Issue |