Vall-E has been trained on over 60,000 hours of speech A Microsoft logo is seen in the picture. | Photo Credit: Reuters With a 3-se...
Vall-E has been trained on over 60,000 hours of speech
With a 3-second audio sample, a new AI from Microsoft can mimic your speech.
Microsoft's Vall-E, which pays homage to OpenAI's Dall-E art generator, is a new AI that has been trained on more than 60,000 hours of speech and has been called a "neural codec language model."
The Vall-E example from Microsoft is available on the GitHub page. With just three seconds of speech, Microsoft's Vall-E can imitate a specific voice, letting the user just enter the text they want that voice to say to produce paragraphs or other content.
Microsoft recently announced plans to spend $10 billion on OpenAI, the company that developed the augmented reality assistant ChatGPT. At the end of November 2022, ChatGPT was introduced.
COMMENTS