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.16 ViewsLikes 0
