Presentation | 2015-10-15 A Study on Speaker-Independent Voice Conversion Using Spectral Differential Filter Based on Neural Network Harunori Koike, Takashi Nose, Takahiro Shinozaki, Akinori Ito, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | In this paper, we propose a novel technique for making the speech individuality of an arbitrary source (input) speaker. The proposed technique use neural network (NN) for many-to-one mapping and the NN is trained with the pairs of multiple source speakers and a target speaker. The conversion of speaker individuality of the input speech is conducted by spectral differential filter. In the previous studies of voice conversion, speaker-dependent approach was proposed where parallel speech data of source and target speakers are used for conversion model training. There is also another approach where the conversion model is trained by using speaker adaptation with a small amount of target speaker's speech. Recently, we proposed speaker-independent voice conversion without using a user's speech in the training step. The purpose of this study is to improve the naturalness of the converted speech in the speaker-independent voice conversion. We directly convert the waveform of the input speaker using a filter whose parameters are obtained by the differential of spectral features before and after feature mapping. An advantage is that the direct waveform conversion alleviate the quality reduction caused by the extraction error of fundamental frequency in the conventional technique. We also show that the naturalness is further improved by variance compensation with affine transformation. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | |
Paper # | SP2015-61 |
Date of Issue | 2015-10-08 (SP) |
Conference Information | |
Committee | SP |
---|---|
Conference Date | 2015/10/15(2days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | Kobe Univ. |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | Speech interface, Synthesis, Dialogue, Application system, etc. |
Chair | Kazunori Mano(Shibaura Inst. of Tech.) |
Vice Chair | Norihide Kitaoka(Tokushima Univ.) |
Secretary | Norihide Kitaoka(Tokyo City Univ.) |
Assistant | Takashi Nose(Tohoku Univ.) / Taichi Asami(NTT) |
Paper Information | |
Registration To | Technical Committee on Speech |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | A Study on Speaker-Independent Voice Conversion Using Spectral Differential Filter Based on Neural Network |
Sub Title (in English) | |
Keyword(1) | |
Keyword(2) | |
Keyword(3) | |
Keyword(4) | |
Keyword(5) | |
1st Author's Name | Harunori Koike |
1st Author's Affiliation | Tohoku University(Tohoku Univ.) |
2nd Author's Name | Takashi Nose |
2nd Author's Affiliation | Tohoku University(Tohoku Univ.) |
3rd Author's Name | Takahiro Shinozaki |
3rd Author's Affiliation | Tokyo Institute of Technology(Tokyo Tech) |
4th Author's Name | Akinori Ito |
4th Author's Affiliation | Tohoku University(Tohoku Univ.) |
Date | 2015-10-15 |
Paper # | SP2015-61 |
Volume (vol) | vol.115 |
Number (no) | SP-253 |
Page | pp.pp.13-18(SP), |
#Pages | 6 |
Date of Issue | 2015-10-08 (SP) |