"DeFake Tool Protects Voice Recordings From Cybercriminals"
Ning Zhang, an assistant professor of computer science and engineering at Washington University in St. Louis, was among three winners of the US Federal Trade Commission's (FTC) Voice Cloning Challenge. "DeFake," Zhang's winning project, uses watermarking for voice recordings. The tool adds carefully crafted distortions that are imperceptible to the human ear to recordings, which makes cloning more difficult by removing usable voice samples. DeFake involves applying adversarial Artificial Intelligence (AI). Voice cloning uses pre-existing speech samples to recreate a voice, typically collected from social media and other platforms. The new tool prevents cloning by making criminally synthesized speech sound like other voices rather than the intended victim. This article continues to discuss the new tool that embeds distortions imperceptible to human ears into audio recordings to prevent cybercriminals from cloning them.
Submitted by grigby1