CloneShield: A Framework for Universal Perturbation Against Zero-Shot Voice Cloning
Li, Renyuan, Liang, Zhibo, Zhang, Haichuan, Shi, Tianyu, Cheng, Zhiyuan, Shi, Jia, Yang, Carl, Tang, Mingjie
–arXiv.org Artificial Intelligence
Recent breakthroughs in text-to-speech (TTS) voice cloning have raised serious privacy concerns, allowing highly accurate vocal identity replication from just a few seconds of reference audio, while retaining the speaker's vocal authenticity. In this paper, we introduce CloneShield, a universal time-domain adversarial perturbation framework specifically designed to defend against zero-shot voice cloning. Our method provides protection that is robust across speakers and utterances, without requiring any prior knowledge of the synthesized text. We formulate perturbation generation as a multi-objective optimization problem, and propose Multi-Gradient Descent Algorithm (MGDA) to ensure the robust protection across diverse utterances. To preserve natural auditory perception for users, we decompose the adversarial perturbation via Mel-spectrogram representations and fine-tune it for each sample. This design ensures imperceptibility while maintaining strong degradation effects on zero-shot cloned outputs. Experiments on three state-of-the-art zero-shot TTS systems, five benchmark datasets and evaluations from 60 human listeners demonstrate that our method preserves near-original audio quality in protected inputs (PESQ = 3.90, SRS = 0.93) while substantially degrading both speaker similarity and speech quality in cloned samples (PESQ = 1.07, SRS = 0.08).
arXiv.org Artificial Intelligence
May-27-2025
- Genre:
- Research Report (0.64)
- Industry:
- Information Technology > Security & Privacy (1.00)
- Technology:
- Information Technology > Artificial Intelligence
- Vision (1.00)
- Speech (1.00)
- Natural Language > Large Language Model (1.00)
- Machine Learning > Statistical Learning (0.66)
- Information Technology > Artificial Intelligence