| Presentation | 2021-03-03 An optimal prediction of phoneme under Bayes criterion by weighting multiple hidden Markov models Taishi Yamaoka, Shota Saito, Toshiyasu Matsushima, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | In this paper, we propose a prediction method for prediction problems using a hidden Markov model. Specifically, it is a proposal for phoneme recognition, which is one of the prediction problems. In the previous studies on phoneme recognition using the Hidden Markov Model, the Hidden Markov Model used for prediction is defined as one by a certain criteria. In addition, for the defined Hidden Markov Model, parameters were estimated from the training data, and the phonemes corresponding to the new speech data were predicted using paremters. In this peper, we assume 0-1 loss as the loss function, and formulate the optimum prediction based on Bayesian criterion. In other words, instead of selecting one Hidden Markov Model and estimating its parameters and making predictions using them, we propose a prediction that directly minimizes the probability of error in the prediction. Although this prediction is theoretically optimal, its calculation involves two problems: (i) The complexity of the sum calculation of the state transition series is on the exponential order with respect to the length of the voice. (ii) It is difficult to analytically calculate the integral by the posterior distribution of the parameters of the Hidden Markov Model. In order to solve these problems, in this paper, we apply the Viterbi algorithm for problem (i) and the Variational Bayesian method for problem (ii), and propose a Bayesian semi-optimal algorithm. This algorithm makes predictions by weighted averages of approximate posterior distributions of multiple Hidden Markov Models. By conducting numerical experiments using artificial data, it was confirmed that the proposed method has a smaller false recognition rate than the method of selecting and predicting one model as in the previous research. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Phoneme recognition / Hidden Markov model / Bayes criteria |
| Paper # | EA2020-76,SIP2020-107,SP2020-41 |
| Date of Issue | 2021-02-24 (EA, SIP, SP) |
| Conference Information | |
| Committee | EA / US / SP / SIP / IPSJ-SLP |
|---|---|
| Conference Date | 2021/3/3(2days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Online |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | Speech, Engineering/Electro Acoustics, Signal Processing, Ultrasonics, and Related Topics |
| Chair | Kenichi Furuya(Oita Univ.) / Hikaru Miura(Nihon Univ.) / Hisashi Kawai(NICT) / Kazunori Hayashi(Kyoto Univ.) / 北岡 教英(豊橋技科大) |
| Vice Chair | Yoshinobu Kajikawa(Kansai Univ.) / Kentaro Matsui(NHK) / Jun Kondo(Shizuoka Univ.) / Yoshikazu Koike(Shibaura Inst. of Tech.) / / Yukihiro Bandou(NTT) / Toshihisa Tanaka(Tokyo Univ. Agri.&Tech.) |
| Secretary | Yoshinobu Kajikawa(Univ. of Tokyo) / Kentaro Matsui(NTT) / Jun Kondo(Doshisha Univ.) / Yoshikazu Koike(Tohoku Univ.) / (Univ. of Tokyo) / Yukihiro Bandou(Waseda Univ.) / Toshihisa Tanaka(Hosei Univ.) / (Waseda Univ.) |
| Assistant | Yukou Wakabayashi(Tokyo Metropolitan Univ.) / Tatsuya Komatsu(LINE) / Shinnosuke Hirata(Tokyo Inst. of Tech.) / Yusuke Ijima(NTT) / Yuichi Tanaka(Tokyo Univ. Agri.&Tech.) |
| Paper Information | |
| Registration To | Technical Committee on Engineering Acoustics / Technical Committee on Ultrasonics / Technical Committee on Speech / Technical Committee on Signal Processing / Special Interest Group on Spoken Language Processing |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | An optimal prediction of phoneme under Bayes criterion by weighting multiple hidden Markov models |
| Sub Title (in English) | |
| Keyword(1) | Phoneme recognition |
| Keyword(2) | Hidden Markov model |
| Keyword(3) | Bayes criteria |
| 1st Author's Name | Taishi Yamaoka |
| 1st Author's Affiliation | Waseda University(Waseda Univ.) |
| 2nd Author's Name | Shota Saito |
| 2nd Author's Affiliation | Waseda University(Waseda Univ.) |
| 3rd Author's Name | Toshiyasu Matsushima |
| 3rd Author's Affiliation | Waseda University(Waseda Univ.) |
| Date | 2021-03-03 |
| Paper # | EA2020-76,SIP2020-107,SP2020-41 |
| Volume (vol) | vol.120 |
| Number (no) | EA-397,SIP-398,SP-399 |
| Page | pp.pp.97-102(EA), pp.97-102(SIP), pp.97-102(SP), |
| #Pages | 6 |
| Date of Issue | 2021-02-24 (EA, SIP, SP) |