Nvidia AI research brings science fiction closer to reality

Nvidia AI research brings science fiction closer to reality

We’re excited to bring back Transform 2022 in person on July 19 and virtually from July 20-28. Join leaders in AI and data for in-depth discussions and exciting networking opportunities. Register today!

Could AI teach multiple skills at the same time? Are immersive displays using holography closer to reality than ever? No one can say for sure what the future of artificial intelligence (AI) holds. But one way to get some insight is to watch the research that Nvidia will present to Siggraph 2022which will be held from August 8 to 11.

Nvidia is collaborating with researchers to present 16 papers at Siggraph 2022, covering several research topics that impact the intersection of graphics and AI technologies.

A paper details innovation with reinforcement learning models, made by researchers at the University of Toronto and UC Berkeley, that could help teach multiple AI skills at the same time.

Another looks at new techniques to help build large-scale virtual worlds with instantaneous neural graph primitives. Getting closer to the technologies that we only see in science fiction, there is also holography research it could one day pave the way for a new display technology that will enable immersive telepresence.

“Our goal is to do work that will impact the business,” David Luebke, vice president of graphics research at Nvidia, told VentureBeat. “It’s about solving problems that people don’t already know the answer to and there’s no easy technical solution, so you have to do some research.”

The intersection of search and enterprise AI

The 16 papers that Nvidia helps showcase focus on innovations that impact graphics, which is what Siggraph is all about. Luebke noted, however, that almost all of the research is also relevant to AI use outside of the graphics realm.

“I consider graphics to be one of the most challenging and interesting applications of computation,” Luebke said. “So it’s no surprise that AI is revolutionizing graphics and that graphics provide a real showcase for AI.”

Luebke said the researchers who worked on the reinforcement learning model actually see themselves as more in the field of robotics than graphics. The model has potential applicability to robots as well as any other AI that needs to learn to perform multiple actions.

“The thing about the graphics is that it’s really, really difficult and it’s really, really compelling,” he said. “Siggraph is a place where we present our graphic achievements, but almost everything we do there is also applicable in a broader context.”

Computational holography and the future of telepresence

Throughout the COVID-19 pandemic, individuals and organizations around the world have suddenly become much more familiar with video conferencing technologies like Zoom. There has also been an increasing use of the use of virtual reality headsets, connecting with the emerging concept of the metaverse. Metaverse and telepresence may one day become significantly more immersive.

One of the papers presented by Nvidia at Siggraph concerns a concept known as computational holography. Luebke explained that at a basic level, computational holography is a technique that can construct a three-dimensional scene, where the human eye can focus anywhere in that scene and see the good thing as if it were really there. The research presented at Siggraph details some new approaches to computational holography that could one day lead to VR headsets that are considerably thinner than current options, providing a more immersive and realistic experience.

“It’s been kind of the holy grail for computer graphics for years and years,” Luebke said of the computational holography work. “This research shows that you can use computation, including neural networks and AI, to improve the quality of holographic displays that work and look great.”

Beyond the papers featured at Siggraph, Luebke said Nvidia research is really interested in telepresence innovations.

VentureBeat’s Mission is to be a digital public square for technical decision makers to learn about transformative enterprise technology and conduct transactions. Learn more about membership.