A Comparison Of Deep Learning MOS Predictors For Speech Synthesis Quality

Speech synthesis quality prediction has made remarkable progress with the development of supervised and self-supervised learning (SSL) MOS predictors but some aspects related to the data are still unclear and require further study. In this paper, we evaluate several MOS predictors based on wav2vec 2...

Full description

Saved in:
Bibliographic Details
Main Authors: Ragano, Alessandro, Benetos, Emmanouil, Chinen, Michael, Martinez, Helard Becerra, Reddy, Chandan K A, Skoglund, Jan, Hines, Andrew
Format: Conference Proceeding
Language:English
Subjects:
Correlation
Deep learning
Natural languages
Predictive models
Self-supervised learning
speech quality prediction
speech synthesis
Training
Training data
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Speech synthesis quality prediction has made remarkable progress with the development of supervised and self-supervised learning (SSL) MOS predictors but some aspects related to the data are still unclear and require further study. In this paper, we evaluate several MOS predictors based on wav2vec 2.0 and the NISQA speech quality prediction model to explore the role of the training data, the influence of the system type, and the role of cross-domain features in SSL models. Our evaluation is based on the VoiceMOS challenge dataset. Results show that SSL-based models show the highest correlation and lowest mean squared error compared to supervised models. The key point of this study is that benchmarking the statistical performance of MOS predictors alone is not sufficient to rank models since potential issues hidden in the data could bias the evaluated performances.
ISSN:2688-1454
DOI:10.1109/ISSC59246.2023.10162088