SALT: Distinguishable Speaker Anonymization Through Latent Space Transformation

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 sp

Stats

• citations8
• S2 citations—
• github stars22
• HF likes0
• heat score9.88
• arxiv keylv2023salt

Related papers

• Distinctive And Natural Speaker Anonymization Via Singular Value Transformation-assisted Matrix (2024)8.82
• MUSA: Multi-lingual Speaker Anonymization Via Serial Disentanglement (2024)0.00
• Language-independent Speaker Anonymization Approach Using Self-supervised Pre-trained Models (2022)9.92
• Reprogramming Self-supervised Learning-based Speech Representations For Speaker Anonymization (2023)2.26
• NPU-NTU System For Voice Privacy 2024 Challenge (2024)7.16
• Analyzing Language-independent Speaker Anonymization Framework Under Unseen Conditions (2022)8.09
• Asynchronous Voice Anonymization Using Adversarial Perturbation On Speaker Embedding (2024)7.16
• A Speech Representation Anonymization Framework Via Selective Noise Perturbation (2022)6.34