NVIDIA brings Gen AI to the physical world at CVPR

NVIDIA is also celebrating its win of the Autonomous Grand Challenge for End-to-End Driving at Scale, as well as the creation of NVIDIA’s largest indoor synthetic dataset, created using Omniverse, for CVPR’s AI City Challenge.

NVIDIA’s 57 papers at the show also break ground in the rapidly advancing field of visual generative AI — developing new techniques to create and interpret images, videos and 3D environments. One such paper, JeDi, proposes a new technique that allows users to easily personalise the output of a diffusion model within a couple of seconds using reference images, significantly outperforming existing methods.

Altogether, these announcements illustrate NVIDIA’s commitment to using generative AI as a new strategy to build applications for use cases in physical AI deployments not only for AV development, but also for industrial environments, healthcare, robotics and more.

You can read more about NVIDIA’s announcements on its blog here.