Problem
Research questions and friction points this paper is trying to address.
End-to-end driving
Open-loop/closed-loop gap
Trajectory optimization
Waypoint supervision
Closed-loop performance
Guiding End-to-End Driving Models with Endpoint-Constrained Trajectory Optimization
🤖 AI Summary
This study addresses the physical inconsistency of trajectories between open-loop training and closed-loop execution in end-to-end autonomous driving models, where infeasible intermediate waypoints cause significant performance degradation. To mitigate this issue, we propose ECO, an unsupervised post-processing layer that reconstructs the geometry of intermediate paths through lightweight constrained optimization by anchoring historical trajectories while preserving predicted endpoints. This plug-and-play method requires neither additional training nor high-definition maps. Experimental results demonstrate that ECO substantially improves closed-loop simulation scores across six models, including VaVAM, achieving a 71% improvement for VaVAM and up to a 123% gain on AlpaSim, while ranking first in the HUGSIM benchmark evaluation.
📝 Abstract
End-to-end driving policies are commonly trained through open-loop behavior cloning, yet they must ultimately operate in closed-loop when deployed on a vehicle, creating a fundamental mismatch between training and execution. Beyond the commonly studied effects of covariate shift and causal confusion, we identify a complementary factor for this open-loop/closed-loop gap: waypoint-based supervision and displacement metrics do not ensure that the intermediate trajectory is physically coherent or easy for the controller to track. We observe that these inconsistencies concentrate primarily at intermediate waypoints, while the predicted endpoint remains comparatively reliable. Based on this observation, we introduce Endpoint-Constrained Optimization (ECO), a lightweight postprocessing layer that anchors the trajectory to the vehicle's executed history, preserves the policy's predicted endpoint, and reshapes the intermediate waypoints to improve feasibility. ECO requires no map, privileged simulator state, or additional training, and can be inserted between a broad range of waypoint-emitting policies and their controllers. Across two closed-loop simulators, it improves the aggregate closed-loop score of all six evaluated generative and regression-based driving policies, and the gains tend to increase with how often the base plans violate motion limits. On HUGSIM, ECO improves VaVAM from 18.1 to 31.0 HD-Score (+71%), achieving 1st place on the HUGSIM Closed-Loop Driving Challenge. Similarly, on AlpaSim, ECO increases the scene scores of VaVAM and DiffusionDrive by 123% and 22%, respectively. These results show that for a broad collection of end-to-end driving models, repairing the intermediate geometry of predicted trajectories without changing the policy's predicted endpoint can substantially improve closed-loop performance.
Innovation
Methods, ideas, or system contributions that make the work stand out.
Endpoint-Constrained Optimization
End-to-End Driving
Trajectory Optimization
Closed-Loop Simulation
Postprocessing Layer
🔎 Similar Papers
No similar papers found.
