Wednesday 11 January 2023

zero-shot text to speech synthesis (10. 406)

Via Waxy, we are introduced frighteningly adept AI protocol from Microsoft called VALL·E that can extrapolate and convincingly simulate, imitate any voice with a sample of just a few seconds of audio. The underpinning code is not being published as an acknowledgement of its potential for misuse but there is a demonstration reel of its abilities that’s pretty amazing and are tuned for diversity, emotional maintenance and environment. Eleven benevolent elephants for your elocution.