VoiceMark: Zero-Shot Voice Cloning-Resistant Watermarking Approach Leveraging Speaker-Specific Latents
This addresses the need for tracing unauthorized voice cloning in zero-shot scenarios, which is an incremental advance over existing watermarking techniques.
The paper tackled the problem of watermarking audio to resist zero-shot voice cloning, where existing methods fail, and proposed VoiceMark, which achieved over 95% detection accuracy after synthesis, significantly outperforming prior methods at around 50%.
Voice cloning (VC)-resistant watermarking is an emerging technique for tracing and preventing unauthorized cloning. Existing methods effectively trace traditional VC models by training them on watermarked audio but fail in zero-shot VC scenarios, where models synthesize audio from an audio prompt without training. To address this, we propose VoiceMark, the first zero-shot VC-resistant watermarking method that leverages speaker-specific latents as the watermark carrier, allowing the watermark to transfer through the zero-shot VC process into the synthesized audio. Additionally, we introduce VC-simulated augmentations and VAD-based loss to enhance robustness against distortions. Experiments on multiple zero-shot VC models demonstrate that VoiceMark achieves over 95% accuracy in watermark detection after zero-shot VC synthesis, significantly outperforming existing methods, which only reach around 50%. See our code and demos at: https://huggingface.co/spaces/haiyunli/VoiceMark