PHONOS: PHOnetic Neutralization for Online Streaming Applications

arXiv:2603.27001 · eess.AS, cs.CL, cs.LG · Submitted 2026-03-27 · Read on arXiv

eess.AS, cs.CL, cs.LG

Submitted: 2026-03-27

Updated: 2026-09-22

Comments: The paper is accepted for publication at SLT 2026 workshop

Code: https://github.com/snakers4/silero-vad

Project page: https://anonymousis23.github.io/demos

License: http://creativecommons.org/licenses/by/4.0/

The gist: Speaker anonymization (SA) systems modify timbre while leaving regional or non-native accent cues intact, which is problematic because such cues can reveal a speaker's first-language or geographic

Terminology

Abstract

Speaker anonymization (SA) systems modify timbre while leaving regional or non-native accent cues intact, which is problematic because such cues can reveal a speaker's first-language or geographic background and narrow the anonymity set. To address this issue, we present PHONOS, a streaming module for real-time SA that performs accent neutralization in a privacy sense: reducing accent-origin cues by converting non-native segmental realizations toward a chosen target accent domain. Our approach pre-generates golden speaker utterances that preserve source timbre and rhythm but replace foreign segmentals with native ones using silence-aware DTW alignment and zero-shot voice conversion. These utterances supervise a causal accent translator that maps non-native content tokens to native equivalents with at most 40ms look-ahead, trained using joint cross-entropy and CTC losses. Our evaluations show an 81% reduction in non-native accent confidence, with listening-test accentedness ratings consistent with this shift. PHONOS also moves outputs away from the original speaker in embedding space, suggesting lower linkability under an embedding-based proxy, while running with at most241, ms end-to-end latency on a single GPU.

Sources

Related papers