Article

Ars TechnicaArs Technica on 2023-01-09 23:15

Microsoft’s new AI can simulate anyone’s voice with 3 seconds of audio

Text-to-speech model can preserve speaker's emotional tone and acoustic environment.

Related news