Presentation | 2014-12-16 Voice conversion based on deep neural network with multiple output sub-networks Tetsuya HASHIMOTO, Yousuke KASHIWAGI, Daisuke SAITO, Keikichi HIROSE, Nobuaki MINEMATSU, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | This paper proposes a novel approach to construct a voice conversion (VC) system using deep neural networks (DNN) with the final goal to realize conversion to arbitrary speakers. Although DNN-based VC achieves certain performance improvement as compared to Gaussian Mixture Models (GMM) based VC, it has low flexibility in conversion processes. This is because functions of layers/nodes are difficult to be interpreted. In order to solve this problem, we introduce a new architecture that has one SI trunk and multiple SD branches, which are simultaneously trained by using parallel data from multiple speaker pairs. The objective evaluation demonstrates that the proposed architecture improves the performance of voice conversion. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | deep neural network / voice conversion / multiple-output / sub-network |
Paper # | SP2014-117 |
Date of Issue |
Conference Information | |
Committee | SP |
---|---|
Conference Date | 2014/12/8(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 | Speech (SP) |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | Voice conversion based on deep neural network with multiple output sub-networks |
Sub Title (in English) | |
Keyword(1) | deep neural network |
Keyword(2) | voice conversion |
Keyword(3) | multiple-output |
Keyword(4) | sub-network |
1st Author's Name | Tetsuya HASHIMOTO |
1st Author's Affiliation | Grad. School of Information Science and Technology, The Univ. of Tokyo() |
2nd Author's Name | Yousuke KASHIWAGI |
2nd Author's Affiliation | Grad. School of Engineering, The Univ. of Tokyo |
3rd Author's Name | Daisuke SAITO |
3rd Author's Affiliation | Grad. School of Information Science and Technology, The Univ. of Tokyo |
4th Author's Name | Keikichi HIROSE |
4th Author's Affiliation | Grad. School of Information Science and Technology, The Univ. of Tokyo |
5th Author's Name | Nobuaki MINEMATSU |
5th Author's Affiliation | Grad. School of Engineering, The Univ. of Tokyo |
Date | 2014-12-16 |
Paper # | SP2014-117 |
Volume (vol) | vol.114 |
Number (no) | 365 |
Page | pp.pp.- |
#Pages | 6 |
Date of Issue |