SALT: Distinguishable Speaker Anonymization Through Latent Space Transformation

TL;DR

SALT introduces a novel speaker anonymization method using latent space interpolation and extrapolation, achieving high speaker distinctiveness while maintaining speech quality and intelligibility, especially for out-of-distribution speakers.

Contribution

The paper presents SALT, a new speaker anonymization system leveraging latent space transformation with interpolation and extrapolation techniques for improved diversity and distinctiveness.

Findings

Achieves state-of-the-art speaker distinctiveness metrics.

Maintains speech quality and intelligibility.

Effective for out-of-distribution speakers.

Abstract

Speaker anonymization aims to conceal a speaker's identity without degrading speech quality and intelligibility. Most speaker anonymization systems disentangle the speaker representation from the original speech and achieve anonymization by averaging or modifying the speaker representation. However, the anonymized speech is subject to reduction in pseudo speaker distinctiveness, speech quality and intelligibility for out-of-distribution speaker. To solve this issue, we propose SALT, a Speaker Anonymization system based on Latent space Transformation. Specifically, we extract latent features by a self-supervised feature extractor and randomly sample multiple speakers and their weights, and then interpolate the latent vectors to achieve speaker anonymization. Meanwhile, we explore the extrapolation method to further extend the diversity of pseudo speakers. Experiments on Voice Privacy Challenge dataset show our system achieves a state-of-the-art distinctiveness metric while preserving speech quality and intelligibility. Our code and demo is availible at https://github.com/BakerBunker/SALT .