Sorry, you need to enable JavaScript to visit this website.

facebooktwittermailshare

Cycle-consistent adversarial networks for non-parallel vocal effort based speaking style conversion

Abstract: 

Speaking style conversion (SSC) is the technology of converting natural speech signals from one style to another. In this study, we propose the use of cycle-consistent adversarial networks (CycleGANs) for converting styles with varying vocal effort, and focus on conversion between normal and Lombard styles as a case study of this problem. We propose a parametric approach that uses the Pulse Model in Log domain (PML) vocoder to extract speech features. These features are mapped using the CycleGAN from utterances in the source style to the corresponding features of target speech. Finally, the mapped features are converted to a Lombard speech waveform with the PML. The CycleGAN was compared in subjective listening tests with 2 other standard mapping methods used in conversion, and the CycleGAN was found to have the best performance in terms of speech quality and in terms of the magnitude of the perceptual change between the two styles.

up
0 users have voted:

Paper Details

Authors:
Junichi Yamagishi, Okko Räsänen, Paavo Alku
Submitted On:
8 May 2019 - 4:29am
Short Link:
Type:
Poster
Event:
Presenter's Name:
Shreyas Seshadri
Paper Code:
2994
Document Year:
2019
Cite

Document Files

Seshadri_ICASSP2019_final.pdf

(17)

Subscribe

[1] Junichi Yamagishi, Okko Räsänen, Paavo Alku, "Cycle-consistent adversarial networks for non-parallel vocal effort based speaking style conversion", IEEE SigPort, 2019. [Online]. Available: http://sigport.org/4047. Accessed: Jul. 22, 2019.
@article{4047-19,
url = {http://sigport.org/4047},
author = {Junichi Yamagishi; Okko Räsänen; Paavo Alku },
publisher = {IEEE SigPort},
title = {Cycle-consistent adversarial networks for non-parallel vocal effort based speaking style conversion},
year = {2019} }
TY - EJOUR
T1 - Cycle-consistent adversarial networks for non-parallel vocal effort based speaking style conversion
AU - Junichi Yamagishi; Okko Räsänen; Paavo Alku
PY - 2019
PB - IEEE SigPort
UR - http://sigport.org/4047
ER -
Junichi Yamagishi, Okko Räsänen, Paavo Alku. (2019). Cycle-consistent adversarial networks for non-parallel vocal effort based speaking style conversion. IEEE SigPort. http://sigport.org/4047
Junichi Yamagishi, Okko Räsänen, Paavo Alku, 2019. Cycle-consistent adversarial networks for non-parallel vocal effort based speaking style conversion. Available at: http://sigport.org/4047.
Junichi Yamagishi, Okko Räsänen, Paavo Alku. (2019). "Cycle-consistent adversarial networks for non-parallel vocal effort based speaking style conversion." Web.
1. Junichi Yamagishi, Okko Räsänen, Paavo Alku. Cycle-consistent adversarial networks for non-parallel vocal effort based speaking style conversion [Internet]. IEEE SigPort; 2019. Available from : http://sigport.org/4047