OmniV2X: A Generative Foundation Planner for Efficient End-to-End Cooperative Driving

OmniV2X is a generative foundation model designed for vehicle-to-everything (V2X) cooperative driving, addressing challenges in computational cost, data scarcity, and compliance with messaging standards. It employs a high-capacity generative sequence planner trained with a downstream trajectory generation loss and integrates multi-modal inputs using cross-attention mechanisms. Evaluated on the DAIR-V2X-Seq dataset, OmniV2X achieves state-of-the-art results while requiring less than 10% of the fine-tune V2X dataset and under 1% of communication bandwidth, making it highly efficient for practitioners in cooperative driving applications.