The rise of audio deepfakes: implications and challenges

Original article: LinkedIn Audio deepfakes have become a growing concern as the technology used to create them has rapidly advanced in recent years. Novel approaches, such as the VALL-E Text-To-Speech (TTS) system developed by Microsoft, claim to synthesize voices with only three seconds of audio from the targeted speaker. Microsoft’s VALL-E research paper was published in early January 2023 and was followed in February by a similar piece of work by Google’s SPEAR-TTS. These technologies are part of a wider development called generative AI, whose recent breakthroughs have been making the headlines over the last months and includes image generation (DALL-E, Stable diffusion, Midjourney,…) and text generation (ChatGPT,…). ...

February 20, 2023 · 3 min · Benoit Fauve

Deepfake audio has a tell: Validsoft can spot it

Original article: LinkedIn Deepfake Audio Has A Tell and ValidSoft Can Spot It Have you seen the recent news about voice cloning abuse using software produced by Eleven Labs? Users took the voice-cloning technology to create deepfake audio recordings of the actress Emma Watson reading Mein Kampf by Adolf Hitler and the broadcaster Sir David Attenborough being racist and posted them on the message board 4Chan in the latest abuse of AI technology. ...

February 1, 2023 · 2 min · Benoit Fauve
Google Scholar · Semantic Scholar · DBLP · ORCID