| Presentation | 2005/12/8 Features Parameter Transformation in Speaker Recognition Using Vowels Takashi OSANAI, Kazuhiko OZEKI, |
|---|---|
| PDF Download Page | PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | In this report, we propose a feature parameter transformation method using the distribution of vowel features in order to improve the accuracy of speaker verification using vowels uttered in isolation. The transformation is performed in two stages. In the first stage, we standardize a parameter by subtracting the average, and then dividing by the standard deviation. In the second stage, we normalize the parameter by the norm of each feature vector. We conducted speaker verification experiments using mel-frequency cepstral coefficients and LPC cepstral coefficients, which are widely used for speaker verification. The result of the experiments showed about 3% improvement of the average speaker verification rate by applying the transformation in both cases. The transformation is closely related to polar-coordinates transformation. Therefore, we decomposed a parameter into the radius component and the circumference component, and conducted speaker verification experiments using each component. As a result, it became clear that speaker individuality information is contained in the circumference component more than in the radius component. It was also confir med that speaker individuality is emphasized in the spectral envelope by applying the proposed transformation. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Speaker Recognition / Vowel / Standardization / Normalization / Direction Cosine |
| Paper # | HIP2005-96 |
| Date of Issue |
| Conference Information | |
| Committee | HIP |
|---|---|
| Conference Date | 2005/12/8(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 | Human Information Processing (HIP) |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Features Parameter Transformation in Speaker Recognition Using Vowels |
| Sub Title (in English) | |
| Keyword(1) | Speaker Recognition |
| Keyword(2) | Vowel |
| Keyword(3) | Standardization |
| Keyword(4) | Normalization |
| Keyword(5) | Direction Cosine |
| 1st Author's Name | Takashi OSANAI |
| 1st Author's Affiliation | Third Information Science Section, Department of Fourth Forensic Science, National Research Institute of Police Science:Department of Information and Communication Engineering, Faculty of Electro-Communications, University of Electro-Communications() |
| 2nd Author's Name | Kazuhiko OZEKI |
| 2nd Author's Affiliation | Department of Information and Communication Engineering, Faculty of Electro-Communications, University of Electro-Communications |
| Date | 2005/12/8 |
| Paper # | HIP2005-96 |
| Volume (vol) | vol.105 |
| Number (no) | 479 |
| Page | pp.pp.- |
| #Pages | 6 |
| Date of Issue | |