In this paper, we describe a novel model training method for one-to-many eigenvoice conversion (EVC). One-to-many EVC is a technique for converting a specific source speaker's voice into an arbitrary target speaker's voice. An eigenvoice Gaussian mixture model (EV-GMM) is trained in advance using multiple parallel data sets consisting of utterance-pairs of the source speaker and many pre-stored target speakers. The EV-GMM can be adapted to new target speakers using only a few of their arbitrary utterances by estimating a small number of adaptive parameters. In the adaptation process, several parameters of the EV-GMM to be fixed for different target speakers strongly affect the conversion performance of the adapted model. In order to improve the conversion performance in one-to-many EVC, we propose an adaptive training method of the EV-GMM. In the proposed training method, both the fixed parameters and the adaptive parameters are optimized by maximizing a total likelihood function of the EV-GMMs adapted to individual pre-stored target speakers. We conducted objective and subjective evaluations to demonstrate the effectiveness of the proposed training method. The experimental results show that the proposed adaptive training yields significant quality improvements in the converted speech.
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copy
Yamato OHTANI, Tomoki TODA, Hiroshi SARUWATARI, Kiyohiro SHIKANO, "Adaptive Training for Voice Conversion Based on Eigenvoices" in IEICE TRANSACTIONS on Information,
vol. E93-D, no. 6, pp. 1589-1598, June 2010, doi: 10.1587/transinf.E93.D.1589.
Abstract: In this paper, we describe a novel model training method for one-to-many eigenvoice conversion (EVC). One-to-many EVC is a technique for converting a specific source speaker's voice into an arbitrary target speaker's voice. An eigenvoice Gaussian mixture model (EV-GMM) is trained in advance using multiple parallel data sets consisting of utterance-pairs of the source speaker and many pre-stored target speakers. The EV-GMM can be adapted to new target speakers using only a few of their arbitrary utterances by estimating a small number of adaptive parameters. In the adaptation process, several parameters of the EV-GMM to be fixed for different target speakers strongly affect the conversion performance of the adapted model. In order to improve the conversion performance in one-to-many EVC, we propose an adaptive training method of the EV-GMM. In the proposed training method, both the fixed parameters and the adaptive parameters are optimized by maximizing a total likelihood function of the EV-GMMs adapted to individual pre-stored target speakers. We conducted objective and subjective evaluations to demonstrate the effectiveness of the proposed training method. The experimental results show that the proposed adaptive training yields significant quality improvements in the converted speech.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.E93.D.1589/_p
Copy
@ARTICLE{e93-d_6_1589,
author={Yamato OHTANI, Tomoki TODA, Hiroshi SARUWATARI, Kiyohiro SHIKANO, },
journal={IEICE TRANSACTIONS on Information},
title={Adaptive Training for Voice Conversion Based on Eigenvoices},
year={2010},
volume={E93-D},
number={6},
pages={1589-1598},
abstract={In this paper, we describe a novel model training method for one-to-many eigenvoice conversion (EVC). One-to-many EVC is a technique for converting a specific source speaker's voice into an arbitrary target speaker's voice. An eigenvoice Gaussian mixture model (EV-GMM) is trained in advance using multiple parallel data sets consisting of utterance-pairs of the source speaker and many pre-stored target speakers. The EV-GMM can be adapted to new target speakers using only a few of their arbitrary utterances by estimating a small number of adaptive parameters. In the adaptation process, several parameters of the EV-GMM to be fixed for different target speakers strongly affect the conversion performance of the adapted model. In order to improve the conversion performance in one-to-many EVC, we propose an adaptive training method of the EV-GMM. In the proposed training method, both the fixed parameters and the adaptive parameters are optimized by maximizing a total likelihood function of the EV-GMMs adapted to individual pre-stored target speakers. We conducted objective and subjective evaluations to demonstrate the effectiveness of the proposed training method. The experimental results show that the proposed adaptive training yields significant quality improvements in the converted speech.},
keywords={},
doi={10.1587/transinf.E93.D.1589},
ISSN={1745-1361},
month={June},}
Copy
TY - JOUR
TI - Adaptive Training for Voice Conversion Based on Eigenvoices
T2 - IEICE TRANSACTIONS on Information
SP - 1589
EP - 1598
AU - Yamato OHTANI
AU - Tomoki TODA
AU - Hiroshi SARUWATARI
AU - Kiyohiro SHIKANO
PY - 2010
DO - 10.1587/transinf.E93.D.1589
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E93-D
IS - 6
JA - IEICE TRANSACTIONS on Information
Y1 - June 2010
AB - In this paper, we describe a novel model training method for one-to-many eigenvoice conversion (EVC). One-to-many EVC is a technique for converting a specific source speaker's voice into an arbitrary target speaker's voice. An eigenvoice Gaussian mixture model (EV-GMM) is trained in advance using multiple parallel data sets consisting of utterance-pairs of the source speaker and many pre-stored target speakers. The EV-GMM can be adapted to new target speakers using only a few of their arbitrary utterances by estimating a small number of adaptive parameters. In the adaptation process, several parameters of the EV-GMM to be fixed for different target speakers strongly affect the conversion performance of the adapted model. In order to improve the conversion performance in one-to-many EVC, we propose an adaptive training method of the EV-GMM. In the proposed training method, both the fixed parameters and the adaptive parameters are optimized by maximizing a total likelihood function of the EV-GMMs adapted to individual pre-stored target speakers. We conducted objective and subjective evaluations to demonstrate the effectiveness of the proposed training method. The experimental results show that the proposed adaptive training yields significant quality improvements in the converted speech.
ER -