NVIDIA’s ‘Omniverse Avatar’ Helps Create Intelligent Virtual Agents

NVIDIA has unveiled its latest platform to help create intelligent virtual agents, known equally the Omniverse Avatar. It uses a combination of facial and speech recognition, synthetic speech, and 3D avatar animation to create the perfect virtual bot. The manufacturer said these factors would assist create a wide variety of virtual agents.

CEO Jensen Huang gave out the information at NVIDIA GTC 2021, the visitor'south annual AI briefing. 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 graphic symbol tin can maintain eye contact with the customers and fifty-fifty reply to facial expressions. Interestingly, the AI character can also offer additional information, including details on protein options equally shown in the presentation.

"This volition be useful for smart retail, drive-throughs, and customer service," Huang said of the engineering science. 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 every bit office of this effort.

This is part of NVIDIA's vision for the hereafter of AI

"We believe that virtual worlds will enable the side by side era of innovation. Virtual worlds are going to help united states visualize things that haven't been congenital notwithstanding, and help united states of america come across 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 Projection Maxine volition likewise employ Omniverse Avatar for AI-based video conferences and virtual collaborations. Huang'due south sit-in shows the platform's ability to bring advanced video and audio features to content cosmos and virtual collaboration apps. This includes real-time translation into multiple languages with the same intonation and vocalization.

NVIDIA seemingly wants to bring virtual elements into the real world with Omniverse Avatar. Allowing these virtual elements to collaborate 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'southward too early to tell if these technologies will proceeds prominence. However, not everybody is likely to be on board with the idea of interacting with virtual characters while ordering their nutrient.

Moreover, Huang noted in the presentation that the virtual character has a response time of two seconds. It'south probable that the company will ameliorate this earlier a wider rollout. But if it isn't, this could be a dealbreaker for many. Overall, NVIDIA's time to come looks pretty exciting, although it's unclear how it could be utilized in the existent world.

Check out Huang's keynote accost from the video below: