CEO Jensen Huang gave out the information at NVIDIA GTC 2021, the company’s annual AI conference. In a presentation, Huang showcased some of the capabilities of Omniverse Avatar (via). One implementation involves a digital kiosk with an animated character helping customers navigate the food menu. More importantly, this character can maintain eye contact with the customers and even respond to facial expressions. Interestingly, the AI character can also offer additional information, including details on protein options as shown in the presentation. “This will be useful for smart retail, drive-throughs, and customer service,” Huang said of the technology. Additionally, he offered another demo involving an animated version of himself answering questions on protein production and climate change. Speaking to members of the media, VP of the Omniverse platform at NVIDIA, Richard Kerris said that several new features have been added to Omniverse Enterprise as part of this effort.
This is part of NVIDIA’s vision for the future of AI
“We believe that virtual worlds will enable the next era of innovation. Virtual worlds are going to help us visualize things that haven’t been built yet, and help us see how to operate and maintain factories as digital twins,” Kerris said. As the name indicates, this is based on NVIDIA’s Omniverse platform. The company describes Omniverse as a platform designed for “virtual collaboration and real-time physically accurate simulation.” Further, NVIDIA’s Project Maxine will also utilize Omniverse Avatar for AI-based video conferences and virtual collaborations. Huang’s demonstration shows the platform’s ability to bring advanced video and audio features to content creation and virtual collaboration apps. This includes real-time translation into multiple languages with the same intonation and voice. NVIDIA seemingly wants to bring virtual elements into the real world with Omniverse Avatar. Allowing these virtual elements to interact with real people in the physical world could be a gamechanger. By contrast, Meta’s vision for a metaverse involves taking users inside the virtual world.
Although the presentations are fairly impressive, it’s too early to tell if these technologies will gain prominence. However, not everybody is likely to be on board with the idea of interacting with virtual characters while ordering their food. Moreover, Huang noted in the presentation that the virtual character has a response time of 2 seconds. It’s likely that the company will improve this before a wider rollout. But if it isn’t, this could be a dealbreaker for many. Overall, NVIDIA’s future looks pretty exciting, although it’s unclear how it could be utilized in the real world. Check out Huang’s keynote address from the video below:
