Meta introduces A Speech Capabilities AI Generative Model
Meta Platforms has made a speech-generating artificial intelligence (AI) model that uses generative AI.
Meta said in a news release on Friday, June 16, that the new Voicebox can help with things like editing audio, sampling, and styling.
“Voicebox can produce high-quality audio clips and edit prerecorded audio — like removing car horns or a dog barking — all while preserving the content and style of the audio,” Meta said in the release. “The model is also multilingual and can produce speech in six languages.”
Voicebox can convert text to speech using audio samples as short as two seconds, re-create parts of speech for editing and noise reduction, and use a person’s voice to read text in any of these six languages, according to a news release.
The press statement said that Voicebox can speak English, French, German, Spanish, Polish, and Portuguese.
This new generative AI tool could be used in the future to give virtual assistants and non-player characters in the metaverse voices that sound natural, to let visually impaired people hear written messages read by their friends, to make and edit audio tracks, and to help people communicate in other languages using their own voices, according to a press release.
“Voicebox is an important step forward in our generative AI research, and we look forward to continuing our exploration in the audio space and seeing how other researchers build on our work,” Meta said in the release.
The technology can tell what people are feeling, give tips, and finish whole transactions.
According to the study “How Consumers Want to Live in the Voice Economy,” 61% of consumers already think voice assistants will become as smart and reliable as human assistants, and 41% think this will happen within five years.
Alphabet, which owns Google, and Microsoft have also talked about using creative AI for voice apps.
In April, both businesses talked about how they are developing and rolling out generative AI tools across the enterprise, such as tools that help with content creation, collaboration, and better, more personalized search results.
