You can now clone your voice! OpenAI launches a new “Voice Engine” model that can reproduce sounds from an audio sample Just listen to the original for just 15 seconds!
OpenAI, a leading artificial intelligence or artificial intelligence company from the United States. After launching new models of chatbots such as GPT-4 and “Sora” Geneative AI that can create short videos just by entering text. I went in mid February.
Recently, OpenAI made waves again after launching “Voice Engine,” an AI model that reproduces voices that can create readings out loud. To be like the original sound
According to OpenAI, the Voice Engine project has been in development since the end of 2022 and is scheduled to be used for reading aloud on Chat-GPT.
The ability of Voice Engine is to listen to the original audio. It only takes 15 seconds. After that, the model will process it and convert it into reproduced audio. It is exactly the same as the original text, and can be read aloud according to the text. It can also be used in more than 57 languages. (There's also Thai!)
However, there are only a few companies. Other companies with access to the technology include education technology company Age of Learning, visual storytelling platform HeyGen, health software maker Dimagi, AI communications app maker Livox, and healthcare platform Lifespan.
Written by Jeff Harris, an OpenAI member who developed OpenAI's voice engine or vocal language model, said that the model has received training regarding screening of what information is allowed to be released and the information has been completed to be released to the public and this model will be available to only about 10 developers.
source: the edge