NVIDIA Corporation is a worldwide technology firm based in Santa Clara, California, that was founded in Delaware. It creates graphics processing units for the gaming and professional markets, as well as system on a chip (SoC) units for mobile computing and automotive applications. Its principal GPU line, dubbed “GeForce,” competes directly with Advanced Micro Devices’ “Radeon”-branded GPUs (AMD). With its mobile game consoles Shield Portable, Shield Tablet, and Shield Android TV, as well as its cloud gaming service GeForce Now, Nvidia has increased its position in the gaming sector. Workstations with its professional GPUs are utilised in sectors including architecture, engineering, and construction, media and entertainment, automotive, scientific research, and manufacturing design.
NVIDIA’s AI assistant technology has been slowly improving in recent months, and now it’s evident how everything fits together. Omniverse Avatar (for 3D assistant development) and Riva (custom AI voice production) are two platforms launched by the firm that, when combined, result in astonishingly lifelike virtual personas with little effort — or, in one case, purposely unrealistic virtual personas.
NVIDIA produced an Omniverse Avatar using a lady’s photo and utilised Riva to train the voice based on that woman, convert text to speech, and translate to several languages in one demo. With the exception of a few stiff-sounding translations, the artificial stand-in looks and sounds almost like to the real person, and can even tilt its head while keeping natural-looking eye contact. This might, as you might expect, lead to more relatable virtual assistants at kiosks and online.
Another demonstration, for NVIDIA’s Project Tokkio “talking kiosk” reference software, demonstrates what might happen when you develop a completely fake character. A 3D, ray-traced toy version of CEO Jensun Huang (complete with his characteristic clothing and a Riva-trained voice) uses AI to hold a conversation with actual people about climate change and the importance of proteins in the body. His face and hands are animated by Omniverse systems. Of course, it’s not intended to be completely authentic, but it demonstrates how a 3D virtual assistant may be far more interesting than a voice.
The majority of Maxine’s development kit is already on hand. Riva is now in open beta and will be free to use for “small-scale” projects. Riva Enterprise, which will arrive in early 2022, will be responsible for larger roll-outs. However, you’ll have to wait a little longer for Omniverse Avatar. While the fundamental Omniverse platform is currently in open beta, Avatar is still “in development” with no release date set. Still, this alludes to a future in which an airport or favourite restaurant can provide a helpful assistance that isn’t overly robotic.