LATIDIA · Ciberseguridad

La vulnerabilidad de las marcas de agua de audio neuronales en la mejora del habla

arXiv: 2609.29040v1Tipo de anuncio: cruz Resumen: Las marcas de agua de audio neuronales se implementan cada vez más en los sistemas comerciales de generación de voz para hacer que el habla generada por IA sea rastreable, pero su robustez se ha estudiado

La noticia

arXiv:2609.29040v1 Announce Type: cross Abstract: Neural audio watermarks are increasingly deployed in commercial speech generation systems to make AI-generated speech traceable, yet their robustness has been studied mainly under conventional signal distortions. Since a watermark can be regarded as imperceptible noise added to the speech signal, a natural question is whether speech enhancement (SE), as a denoising model, can remove it. In this paper, we cascade Gaussian noise with SE models as a black-box watermark removal attack, covering both discriminative and generative SE paradigms, against six neural watermarks: AudioSeal, WavMark, SilentCipher, Timbre, Perth, and AlignMark. Experimental results show that the proposed attack significantly outperforms existing neural re-synthesis methods in watermark