New voice cloning AI lets \”you\” speak multiple languages
This article is part of Future Explored – a weekly guide on the latest technology. Subscribe here to receive stories like this in your inbox each Thursday morning.
Microsoft announced in January that it had developed an AI which can clone the voice of a speaker after only hearing them speak for three seconds. VALL-E was not the first AI to clone voices, but its accuracy and the need for a very small audio sample raised the bar.
Microsoft has raised the bar with a new update called \”VALLE X.\” It can clone voices from short samples (4-10 seconds), and use them to synthesize spoken language in another language while maintaining the speaker’s original voice, tone, and emotion.