Super Agents and Confounders: Influence of surrounding agents on vehicle trajectory prediction

2026-04-03

AI Summary

This work addresses the performance degradation of trajectory prediction models in highly interactive driving scenarios, where models often misuse information from surrounding traffic agents, leading to non-causal and unstable decision mechanisms. The study systematically identifies neighboring agents as confounding factors—a previously overlooked issue—and introduces an explainability analysis based on Shapley values to quantify non-robust dependencies on contextual information. To mitigate this, the authors propose a conditional information bottleneck (CIB) mechanism that adaptively compresses and filters uninformative features without requiring additional supervision. Extensive experiments demonstrate that CIB consistently enhances both prediction accuracy and robustness to perturbations across multiple datasets and model architectures.

Abstract

In highly interactive driving scenes, trajectory prediction is conditioned on information from surrounding traffic participants such as cars and pedestrians. Our main contribution is a comprehensive analysis of state-of-the-art trajectory predictors, which reveals a surprising and critical flaw: many surrounding agents degrade prediction accuracy rather than improve it. Using Shapley-based attribution, we rigorously demonstrate that models learn unstable and non-causal decision-making schemes that vary significantly across training runs. Building on these insights, we propose to integrate a Conditional Information Bottleneck (CIB), which does not require additional supervision and is trained to effectively compress agent features as well as ignore those that are not beneficial for the prediction task. Comprehensive experiments using multiple datasets and model architectures demonstrate that this simple yet effective approach not only improves overall trajectory prediction performance in many cases but also increases robustness to different perturbations. Our results highlight the importance of selectively integrating contextual information, which can often contain spurious or misleading signals, in trajectory prediction. Moreover, we provide interpretable metrics for identifying non-robust behavior and present a promising avenue towards a solution.

Problem

• trajectory prediction
• surrounding agents
• confounders
• non-causal decision-making
• spurious signals

Innovation

• Conditional Information Bottleneck
• trajectory prediction
• Shapley attribution
• robustness
• spurious correlation