Boosting Star-GANs for Voice Conversion with Contrastive Discriminator

TL;DR

This paper introduces SimSiam-StarGAN-VC, a novel voice conversion framework that enhances training stability and prevents overfitting by integrating contrastive learning with a Siamese discriminator, leading to superior performance.

Contribution

It proposes a new contrastive learning-based discriminator for StarGAN-VC, improving training stability and conversion quality in nonparallel multi-domain voice conversion.

Findings

Outperforms existing methods on VCC 2018 dataset

Improves training stability and prevents overfitting

Achieves higher objective and subjective quality metrics

Abstract

Nonparallel multi-domain voice conversion methods such as the StarGAN-VCs have been widely applied in many scenarios. However, the training of these models usually poses a challenge due to their complicated adversarial network architectures. To address this, in this work we leverage the state-of-the-art contrastive learning techniques and incorporate an efficient Siamese network structure into the StarGAN discriminator. Our method is called SimSiam-StarGAN-VC and it boosts the training stability and effectively prevents the discriminator overfitting issue in the training process. We conduct experiments on the Voice Conversion Challenge (VCC 2018) dataset, plus a user study to validate the performance of our framework. Our experimental results show that SimSiam-StarGAN-VC significantly outperforms existing StarGAN-VC methods in terms of both the objective and subjective metrics.