Speaker Anonymization Using Neural Audio Codec Language Models

Abstract

The vast majority of approaches to speaker anonymization involve the extraction of fundamental frequency estimates, linguistic features and a speaker embedding which is perturbed to obfuscate the speaker identity before an anonymized speech waveform is resynthesized using a vocoder. Recent work has shown that x-vector transformations are difficult to control consistently: other sources of speaker information contained within fundamental frequency and linguistic features are re-entangled upon vocoding, meaning that anonymized speech signals still contain speaker information. We propose an approach based upon neural audio codecs (NACs), which are known to generate high-quality synthetic speech when combined with language models. NACs use quantized codes, which are known to effectively bottleneck speaker-related information: we demonstrate the potential of speaker anonymization systems based on NAC language modeling by applying the evaluation framework of the Voice Privacy Challenge 2022.

Stats

• citations28
• S2 citations—
• github stars0
• HF likes0
• heat score10.97
• arxiv keypanariello2023speaker

Related papers

• Speaker Anonymization Using X-vector And Neural Waveform Models (2019)0.00
• NPU-NTU System For Voice Privacy 2024 Challenge (2024)7.16
• Language-independent Speaker Anonymization Approach Using Self-supervised Pre-trained Models (2022)9.92
• Voiceprivacy 2022 System Description: Speaker Anonymization With Feature-matched F0 Trajectories (2022)0.00
• Analyzing Language-independent Speaker Anonymization Framework Under Unseen Conditions (2022)8.09
• Speaker Anonymization With Distribution-preserving X-vector Generation For The Voiceprivacy Challenge 2020 (2020)0.00
• Reprogramming Self-supervised Learning-based Speech Representations For Speaker Anonymization (2023)2.26
• Exploring The Importance Of F0 Trajectories For Speaker Anonymization Using X-vectors And Neural Waveform Models (2021)0.00