D2-V2X: Depth-Driven Cooperative V2X Reasoning for Autonomous Driving
For autonomous driving systems, this work addresses sensor occlusion by grounding VLMs in cooperative V2X data, though the 53.5 F1-score and identified 3D-to-2D projection bottleneck indicate incremental progress.
D2-V2X introduces a spatially-aware V2X benchmark with 8,500 QRA triplets and a baseline that aligns LiDAR features with VLM latent space, achieving 24.4% recall in occluded hazard detection (vs near-zero for zero-shot) and 77% reduction in spatial estimation error for visible objects.
Single-vehicle Vision-Language Models (VLMs) are fundamentally constrained by sensor occlusions. While Vehicle-to-Everything (V2X) systems mitigate this, current benchmarks lack the cooperative reasoning required for resolving ambiguities in complex environments. We introduce D2-V2X, a spatially-aware Question-Rationale-Answer (QRA) benchmark featuring 8,500 triplets derived from multimodal vehicle and infrastructure sensors. We additionally establish a baseline that aligns 3D LiDAR features with the VLM's latent space. By enforcing natural language Chain-of-Thought rationales prior to structured JSON outputs, our model is forced to explicitly articulate spatial relations. Our experiments demonstrate that grounding VLMs in cooperative LiDAR achieves 24.4% recall in identifying occluded hazards compared to near-zero in zero-shot models and reduces spatial estimation error for visible objects by 77% compared to the zero-shot baseline. While the model achieves a functional decision-making F1-score of 53.5, we identify 3D-to-2D projection as a fundamental bottleneck in current VLM architectures, establishing a new baseline for future innovation. Data, code, and trained models available at https://github.com/KevinRichard1/D2-V2X