Voice Conversion Based on STRAIGHT and UBM-GMM

被引：0

作者：

Gao Yingying ^{[1
]}

Zhu Weibin ^{[1
]}

机构：

[1] Beijing Jiaotong Univ, Inst Informat Sci, Beijing, Peoples R China

来源：

PROCEEDINGS OF 2009 CONFERENCE ON COMMUNICATION FACULTY | 2009年

关键词：

voice conversion; STRAIGHT; UBM-GMM; TRANSFORMATION;

D O I：

暂无

中图分类号：

TM [电工技术]; TN [电子技术、通信技术];

学科分类号：

0808 ; 0809 ;

摘要：

A novel approach for voice conversion is proposed which uses STRAIGHT analysis/synthesis arithmetic to analyze and synthesize speech. UBM-GMM is adopted and the speaker characters are described by the parameters of GMM. Consequently the amount of the target speaker training dataset is reduced. The transformation function based on GMM is also modified and the parallel corpus is successfully avoided then, along with an improvement in the flexibility of the function to the character parameters.

引用

页码：342 / 345

页数：4

共 8 条

[1] ABE M, 1988, P ICASSP, P655
[2] [Anonymous], 2003, 8 EUR C SPEECH COMM
[3] Restructuring speech representations using a pitch-adaptive time-frequency smoothing and an instantaneous-frequency-based F0 extraction:: Possible role of a repetitive structure in sounds
Kawahara, H
Masuda-Katsuse, I
de Cheveigné, A
[J]. SPEECH COMMUNICATION, 1999, 27 (3-4) : 187 - 207
[4] TRANSFORMATION OF FORMANTS FOR VOICE CONVERSION USING ARTIFICIAL NEURAL NETWORKS
NARENDRANATH, M
MURTHY, HA
RAJENDRAN, S
YEGNANARAYANA, B
[J]. SPEECH COMMUNICATION, 1995, 16 (02) : 207 - 216
[5] SPEAKER IDENTIFICATION AND VERIFICATION USING GAUSSIAN MIXTURE SPEAKER MODELS
REYNOLDS, DA
[J]. SPEECH COMMUNICATION, 1995, 17 (1-2) : 91 - 108
[6] SHUANG ZW, P ICASSP 2008, P4661
[7] Continuous probabilistic transform for voice conversion
Stylianou, Y
Cappe, O
Moulines, E
[J]. IEEE TRANSACTIONS ON SPEECH AND AUDIO PROCESSING, 1998, 6 (02): : 131 - 142
[8] VOICE TRANSFORMATION USING PSOLA TECHNIQUE
VALBRET, H
MOULINES, E
TUBACH, JP
[J]. SPEECH COMMUNICATION, 1992, 11 (2-3) : 175 - 187

← 1 →