Nvidia is trying to make it easier to add conversational capabilities to its platforms with the release of its new Jarvis framework. The platform has been around in a preview form since May of 2020, but will now be made available to the general public.
The Jarvis framework comes with a collection of software tools and AI models trained with more than 1 billion pages of text and 60,000 hours of speech data. Developers can supplement that with custom data to create models that are optimized for a specific task, while the resulting solutions can be deployed through the cloud or embedded in a data center or an edge device. The platform itself supports a number of different voice utilities, including real-time translation and text-to-speech in addition to basic speech recognition.
Nvidia is hoping that Jarvis will help meet the growing demand for voice technologies. Allied Market Research has predicted that the speech recognition market will hit $29.28 billion by 2026, while Mordor Intelligence believes that the chatbot market will climb to $102.29 billion in the same timeframe. In that regard, NewVoiceMedia has reported that 25 percent of the public would already prefer to go through a chatbot for assistance when dealing with a brand.
In other news, Nvidia has formed a partnership with Mozilla Common Voice in an effort to improve the Jarvis platform moving forward. Common Voice is an open source, public domain data set that contains more than 9,000 hours of voice in 60 different languages. Nvidia plans to use that information to train new voice apps and services, and indicated that it will offer those apps for free to the developer community once they are complete.
“We launched Common Voice to teach machines how real people speak in their unique languages, accents, and speech patterns,” said Mozilla Executive Director Mark Surman. “Nvidia and Mozilla have a common vision of democratizing voice technology — and ensuring that it reflects the rich diversity of people and voices that make up the internet.”
Nvidia will update Jarvis with new features through its open beta program in the second quarter of 2021. The company was also one of the organizations that backed the speech recognition specialist Deepgram in a recent $12 million round of Series A funding.