NAVSIM

The 'NAVSIM' dataset/benchmark contains driving scenarios designed to evaluate the performance of autonomous driving systems in understanding and reasoning about complex spatial-temporal relationships in 3D environments.

Papers using NAVSIM (6)

Drive-JEPA: Video JEPA Meets Multimodal Trajectory Distillation for End-to-End Driving2026DriveMA: Driving Vision-Language-Action Models with verifiable Meta-Actions2026DriveMA: Rethinking Language Interfaces in Driving VLAs with One-Step Meta-Actions2026NoRD: A Data-Efficient Vision-Language-Action Model that Drives without Reasoning2026FLARE: Learning Future-Aware Latent Representations from Vision-Language Models for Autonomous Driving2026SGDrive: Scene-to-Goal Hierarchical World Cognition for Autonomous Driving2026