{"id":133495,"date":"2025-06-11T11:00:18","date_gmt":"2025-06-11T11:00:18","guid":{"rendered":"https:\/\/blogs.nvidia.com\/?p=81930"},"modified":"2025-06-11T11:00:18","modified_gmt":"2025-06-11T11:00:18","slug":"nvidia-scores-consecutive-win-for-end-to-end-autonomous-driving-grand-challenge-at-cvpr-2","status":"publish","type":"post","link":"http:\/\/webdomino.net\/index.php\/nvidia\/nvidia-scores-consecutive-win-for-end-to-end-autonomous-driving-grand-challenge-at-cvpr-2\/","title":{"rendered":"NVIDIA Scores Consecutive Win for End-to-End Autonomous Driving Grand Challenge at CVPR"},"content":{"rendered":"
<\/div>

NVIDIA was today named an Autonomous Grand Challenge<\/a> winner at the Computer Vision and Pattern Recognition<\/a> (CVPR) conference, held this week in Nashville, Tennessee. The announcement was made at the Embodied Intelligence for Autonomous Systems on the Horizon Workshop<\/a>.<\/p>\n

This marks the second consecutive year that NVIDIA\u2019s topped the leaderboard in the End-to-End Driving at Scale category and the third year in a row winning an Autonomous Grand Challenge award at CVPR.<\/p>\n

The theme of this year\u2019s challenge was \u201cTowards Generalizable Embodied Systems\u201d \u2014 based on NAVSIM v2<\/a>, a data-driven, nonreactive autonomous vehicle (AV<\/a>) simulation framework.<\/p>\n

The challenge offered researchers the opportunity to explore ways to handle unexpected situations, beyond using only real-world human driving data, to accelerate the development of smarter, safer AVs.<\/p>\n

Generating Safe and Adaptive Driving Trajectories<\/b><\/h2>\n

Participants of the challenge were tasked with generating driving trajectories from multi-sensor data in a semi-reactive simulation, where the ego vehicle\u2019s plan is fixed at the start, but background traffic changes dynamically.<\/p>\n

Submissions were evaluated using the Extended Predictive Driver Model Score, which measures safety, comfort, compliance and generalization across real-world and synthetic scenarios \u2014 pushing the boundaries of robust and generalizable autonomous driving research.<\/p>\n

The NVIDIA AV Applied Research Team<\/a>\u2019s key innovation was the Generalized Trajectory Scoring (GTRS) method, which generates a variety of trajectories and progressively filters out the best one.<\/p>\n

\"\"
GTRS model architecture showing a unified system for generating and scoring diverse driving trajectories using diffusion- and vocabulary-based trajectories.<\/figcaption><\/figure>\n

GTRS introduces a combination of coarse sets of trajectories covering a wide range of situations and fine-grained trajectories for safety-critical situations, created using a diffusion policy conditioned on the environment. GTRS then uses a transformer decoder distilled from perception-dependent metrics, focusing on safety, comfort and traffic rule compliance. This decoder progressively filters out the most promising trajectory candidates by capturing subtle but critical differences between similar trajectories.<\/p>\n

This system has proved to generalize well to a wide range of scenarios, achieving state-of-the-art results on challenging benchmarks and enabling robust, adaptive trajectory selection in diverse and challenging driving conditions.<\/p>\n