WavMark: Watermarking for Audio Generation
This provides a proactive defense mechanism for audio generation technologies, addressing security concerns in voice synthesis and copyright protection.
The paper tackles the risks of voice fraud and impersonation in zero-shot voice synthesis by introducing an audio watermarking framework that encodes up to 32 bits of watermark in a 1-second audio snippet, achieving an average Bit Error Rate of 0.48% across ten common attacks, a reduction of over 2800% compared to state-of-the-art methods.
Recent breakthroughs in zero-shot voice synthesis have enabled imitating a speaker's voice using just a few seconds of recording while maintaining a high level of realism. Alongside its potential benefits, this powerful technology introduces notable risks, including voice fraud and speaker impersonation. Unlike the conventional approach of solely relying on passive methods for detecting synthetic data, watermarking presents a proactive and robust defence mechanism against these looming risks. This paper introduces an innovative audio watermarking framework that encodes up to 32 bits of watermark within a mere 1-second audio snippet. The watermark is imperceptible to human senses and exhibits strong resilience against various attacks. It can serve as an effective identifier for synthesized voices and holds potential for broader applications in audio copyright protection. Moreover, this framework boasts high flexibility, allowing for the combination of multiple watermark segments to achieve heightened robustness and expanded capacity. Utilizing 10 to 20-second audio as the host, our approach demonstrates an average Bit Error Rate (BER) of 0.48\% across ten common attacks, a remarkable reduction of over 2800\% in BER compared to the state-of-the-art watermarking tool. See https://aka.ms/wavmark for demos of our work.