Distinguishable Speaker Anonymization based on Formant and Fundamental Frequency Scaling

TL;DR

This paper introduces a speaker anonymization method that scales formant and fundamental frequency to better preserve speaker identity while protecting privacy, outperforming previous approaches.

Contribution

The paper proposes a novel formant and F0 scaling approach for speaker anonymization that maintains speaker distinguishability and offers adjustable privacy-utility trade-offs.

Findings

Improved speaker distinguishability in anonymized speech.

Significant outperformance over previous frameworks.

Ability to balance privacy and utility with scaling factors.

Abstract

Speech data on the Internet are proliferating exponentially because of the emergence of social media, and the sharing of such personal data raises obvious security and privacy concerns. One solution to mitigate these concerns involves concealing speaker identities before sharing speech data, also referred to as speaker anonymization. In our previous work, we have developed an automatic speaker verification (ASV)-model-free anonymization framework to protect speaker privacy while preserving speech intelligibility. Although the framework ranked first place in VoicePrivacy 2022 challenge, the anonymization was imperfect, since the speaker distinguishability of the anonymized speech was deteriorated. To address this issue, in this paper, we directly model the formant distribution and fundamental frequency (F0) to represent speaker identity and anonymize the source speech by the uniformly scaling formant and F0. By directly scaling the formant and F0, the speaker distinguishability degradation of the anonymized speech caused by the introduction of other speakers is prevented. The experimental results demonstrate that our proposed framework can improve the speaker distinguishability and significantly outperforms our previous framework in voice distinctiveness. Furthermore, our proposed method also can trade off the privacy-utility by using different scaling factors.