Improving Voice Conversion for Dissimilar Speakers Using Perceptual Losses

TL;DR

This paper addresses improving voice conversion techniques for dissimilar speakers by incorporating perceptual losses to enhance the naturalness and similarity of converted speech, aiming to better anonymize speaker identity while maintaining intelligibility.

Contribution

It introduces a novel voice conversion method utilizing perceptual losses to improve conversion quality for dissimilar speakers, advancing privacy-preserving speech processing.

Findings

Enhanced voice conversion quality demonstrated in experiments

Improved speaker similarity metrics achieved

Better anonymization of speaker identity

Abstract

The rising trend of using voice as a means of interacting with smart devices has sparked worries over the protection of users' privacy and data security. These concerns have become more pressing, especially after the European Union's adoption of the General Data Protection Regulation (GDPR). The information contained in an utterance encompasses critical personal details about the speaker, such as their age, gender, socio-cultural origins and more. If there is a security breach and the data is compromised, attackers may utilise the speech data to circumvent the speaker verification systems or imitate authorised users. Therefore, it is pertinent to anonymise the speech data before being shared across devices, such that the source speaker of the utterance cannot be traced. Voice conversion (VC) can be used to achieve speech anonymisation, which involves altering the speaker's characteristics while preserving the linguistic content.