NAVSIMv-2
Emerging7papers using it
2025first seen
NAVSIMv-2 is a dataset/benchmark designed to evaluate the performance of Vision-Language-Action driving models by providing a comprehensive set of driving scenarios and associated multimodal data.
Papers using NAVSIMv-2 (7)
- DriveStack-VLA: Render-Teacher Alignment for BEV-Based DeepStack Vision-Language-Action ModelLaST-VLA: Thinking in Latent Spatio-Temporal Space for Vision-Language-Action in Autonomous DrivingDriveWorld-VLA: Unified Latent-Space World Modeling with Vision-Language-Action for Autonomous DrivingHiST-VLA: A Hierarchical Spatio-Temporal Vision-Language-Action Model for End-to-End Autonomous DrivingDriveFine: Refining-Augmented Masked Diffusion VLA for Precise and Robust DrivingMindDrive: An All-in-One Framework Bridging World Models and Vision-Language Model for End-to-End Autonomous DrivingIRL-VLA: Training an Vision-Language-Action Policy via Reward World Model