Search papers, labs, and topics across Lattice.
This paper introduces a query-based cost-learning framework for end-to-end driving that estimates bounded costs over dynamically reachable ego trajectory queries, addressing limitations of waypoint regression in adapting to safety constraints. The approach leverages compact joint scene tokens for multimodal agent futures and employs contingency-aware cost aggregation with cost-guided MPPI mixing to generate feasible plans. Experimental results on nuScenes and real-world driving logs demonstrate improved collision rates over prior methods while maintaining competitive trajectory quality and an interpretable cost interface.
End-to-end driving models can learn interpretable safety costs over reachable futures, drastically reducing collisions without fine-tuning.
End-to-end planners based on waypoint regression achieve strong open-loop accuracy, but they primarily learn to mimic expert geometry and remain difficult to adapt to deployment-time safety constraints. We propose a query-based cost-learning framework that estimates bounded costs for dynamically reachable ego trajectory queries, rather than dense BEV cells or a small regressed trajectory set. Compact joint scene tokens capture coherent multimodal agent futures, while contingency-aware cost aggregation and cost-guided intra-cluster MPPI mixing convert the learned cost topology into feasible ego plans. On nuScenes, our method improves over prior cost-estimation planners such as ST-P3 and NMP, outperforms most regression baselines in collision rate, while remaining competitive in L2, and retaining an interpretable cost interface. On real-world driving logs, the proposed planner reduces collision rates compared with SparseDrive and Alpamayo without fine-tuning, while maintaining a diverse set of candidate trajectories.
