Causality-Aware End-to-End Autonomous Driving via Ego-Centric Joint Scene Modeling

2026-05-13

AI Summary

This work addresses the lack of explicit modeling of causal interactions between the ego vehicle and surrounding traffic participants in existing end-to-end autonomous driving approaches, which often leads to inconsistent planning in highly interactive scenarios. To this end, we propose CaAD, a novel framework that introduces, for the first time in an end-to-end system, an ego-centric joint causal modeling paradigm. This paradigm captures bidirectional causal dependencies within a shared latent scene representation and incorporates a causal-aware policy alignment mechanism based on joint modality embeddings to ensure consistent and reliable closed-loop planning. Evaluated on Bench2Drive, our method achieves a driving score of 87.53 and a success rate of 71.81%, while attaining a PDMS of 91.1 on NAVSIM, significantly outperforming current state-of-the-art approaches.

Abstract

End-to-end autonomous driving, which bypasses traditional modular pipelines by directly predicting future trajectories from sensor inputs, has recently achieved substantial progress. However, existing methods often overlook the causal inter-dependencies in ego-vehicle planning, ignoring the reciprocal relations between the ego vehicle and surrounding agents. This causal oversight leads to inconsistent and unreliable trajectory predictions, especially in interaction-critical scenarios where ego decisions and neighboring agent behaviors must be reasoned about jointly. To address this limitation, we propose CaAD, a Causality-aware end-to-end Autonomous Driving framework that captures these dependencies within a shared latent scene representation. First, we propose a ego-centric joint-causal modeling module that builds on the marginal prediction branch, and learns causal dependencies between the ego vehicle and interaction-relevant agents. Second, we employ a causality-aware policy alignment stage implemented with joint-mode embeddings to align the stochastic ego policy with planning-oriented closed-loop feedback computed from surrounding traffic and map context. On the Bench2Drive and NAVSIM benchmarks, CaAD demonstrates strong closed-loop planning performance, achieving a Driving Score of 87.53 and Success Rate of 71.81 on Bench2Drive, and a PDMS of 91.1 on NAVSIM.

Problem

• causality
• end-to-end autonomous driving
• ego-vehicle planning
• interactive trajectory prediction
• scene modeling

Innovation

• causality-aware
• end-to-end autonomous driving
• ego-centric modeling
• joint scene representation
• policy alignment