Spotify tests AI's new feature: cloning the voice of a podcast anchor and translating it into other languages

Thank CTOnews.com netizens W (U) for the clue delivery! CTOnews.com, September 25 (Xinhua)-- Spotify is testing a new artificial intelligence feature called Voice Translation (Voice Translation), which can clone a podcast's voice and use it to read podcasts in another language.

The tool is developed by Spotify with the help of OpenAI's automatic speech recognition (ASR) system Whisper. It uses a voice-to-text generation AI model to translate audio files, and a speech copy model to match the style of the original speaker.

CTOnews.com noted that the first hosts to participate in the new feature included Dex Shepard, Monica Padman, Lex Friedman, Bill Simmons and Steven Bartlett. However, not every episode of their respective podcast is immediately available in multiple languages, and currently only Spanish translations are available.

Spotify said it would release more episodes in the coming days and weeks, as well as French and German translations.