Goal-Oriented Reactive Simulation for Closed-Loop Trajectory Prediction

TL;DR

This paper introduces a goal-oriented, reactive simulation framework for closed-loop trajectory prediction that improves collision avoidance by training models in an on-policy setting with self-induced states.

Contribution

It presents a novel transformer-based scene decoder and a hybrid simulation approach for more realistic, reactive, and effective trajectory prediction training.

Findings

Closed-loop training reduces collision rates by up to 27% on nuScenes.

Reactive simulation improves collision avoidance in dense intersections by 79.5%.

Hybrid simulation balances interactivity and stability effectively.

Abstract

Current trajectory prediction models are primarily trained in an open-loop manner, which often leads to covariate shift and compounding errors when deployed in real-world, closed-loop settings. Furthermore, relying on static datasets or non-reactive log-replay simulators severs the interactive loop, preventing the ego agent from learning to actively negotiate surrounding traffic. In this work, we propose an on-policy closed-loop training paradigm optimized for high-frequency, receding horizon ego prediction. To ground the ego prediction in a realistic representation of traffic interactions and to achieve reactive consistency, we introduce a goal-oriented, transformer-based scene decoder, resulting in an inherently reactive training simulation. By exposing the ego agent to a mixture of open-loop data and simulated, self-induced states, the model learns recovery behaviors to correct its own execution errors. Extensive evaluation demonstrates that closed-loop training significantly enhances collision avoidance capabilities at high replanning frequencies, yielding relative collision rate reductions of up to 27.0% on nuScenes and 79.5% in dense DeepScenario intersections compared to open-loop baselines. Additionally, we show that a hybrid simulation combining reactive with non-reactive surrounding agents achieves optimal balance between immediate interactivity and long-term behavioral stability.