Waymo

Waymo is a dataset that contains annotated real-world traffic scenarios used to evaluate the spatial reasoning and scene understanding capabilities of Vision Language Models through Visual Question Answering.

Papers using Waymo (8)

Embodied Scene Understanding for Vision Language Models via MetaVQA2025 · 2 citesNoRD: A Data-Efficient Vision-Language-Action Model that Drives without Reasoning2026Goal-based Vision-language Driving2025AutoVLA: A Vision-Language-Action Model for End-to-End Autonomous Driving with Adaptive Reasoning and Reinforcement Fine-Tuning2025Nurisk: A Visual Question Answering Dataset For Agent-level Risk Assessment In Autonomous Driving2025Nav-ee: Navigation-guided Early Exiting For Efficient Vision-language Models In Autonomous Driving2025AD-EE: Early Exiting For Fast And Reliable Vision-language Models In Autonomous Driving2025MMHU: A Massive-scale Multimodal Benchmark For Human Behavior Understanding2025