Recently, OpenAI introduced the Artificial Intelligence (AI)-powered bot ChatGPT for public testing and it just took over the internet by storm.
People are in awe of ChatGPT's smart repartee to queries and many are calling it a big leap in technology. However, it has spooked Google into overdrive mode to improve its search engine before ChatGPT could do any damage to its business.
Impressed with ChatGPT's prowess, Microsoft is contemplating to invests a whopping $10 billion in Sam Altman-led company OpenAI. There are reports of the company implementing ChatGPT's tech in its newer version of its own Edge browser and Bing search applications.
Now, the Redmond-based company has announced VALL-E, an AI-based neural codec language model that can replicate anybody's voice, and what's more interesting is that it needs just three seconds of audio recording.
Just give a text, and it will speak in a particular person's voice with ease. It is said that VALL-E's speech can match the timbre and even emotional tone of the speaker, and also the acoustic environment of a room too.
"During the pre-training stage, we scale up the TTS (Text-To-Speech synthesis) training data to 60,000 hours of English speech which is hundreds of times larger than existing systems. VALL-E emerges in-context learning capabilities and can be used to synthesize high-quality personalized speech with only a 3-second enrolled recording of an unseen speaker as an acoustic prompt," reads Microsoft's research paper.
For now, Microsoft will not be making the VALL-E's source code public just yet, as it may be misused.
Even the research paper on ChatGPT has revealed that people may use it to spread propaganda and disinformation.
Microsoft will further test the VALL-E more rigorously to ensure, there is less scope for misuse.
So, don't hold your breath. It will be a while before VALL-E makes it to the public domain.
Get the latest news on new launches, gadget reviews, apps, cybersecurity, and more on personal technology only on DH Tech.