From NVIDIA: 2024-06-17 09:00:10

NVIDIA wins Autonomous Grand Challenge at CVPR conference in Seattle, showcasing Hydra-MDP model for End-to-End Driving. Generative AI plays key role in autonomous vehicle development. NVIDIA also introduces Omniverse Cloud Sensor RTX to accelerate autonomous machines’ development.

End-to-End Driving involves training model on AI supercomputer, testing in simulation, and processing sensor data in real time through DRIVE AGX platform. Unified model streamlines AV software development for holistic approach. Hydra-MDP model demonstrates success in handling real-world scenarios for autonomous driving.

CVPR challenge focuses on end-to-end AV model development using nuPlan dataset. NVIDIA Research’s winning model generates safe vehicle path for post-sensor input scenarios. Workflow can be replicated in high-fidelity simulated environments with Omni verse. NVIDIA ranks second for Driving with Language approach in autonomous driving systems.

Over 50 NVIDIA papers accepted at CVPR event, covering topics including automotive, healthcare, and robotics. Sanja Fidler to speak on vision language models at CVPR workshop. Learn more about NVIDIA Research team working on AI, computer graphics, computer vision, self-driving cars, and robotics.



Read more at NVIDIA: NVIDIA Research Wins CVPR Autonomous Grand Challenge