用因果干预消除自动驾驶模型的偏见，提升复杂场景下的安全性和可靠性。
CausalVAD: De-confounding End-to-End Autonomous Driving via Causal Intervention
- 设计轻量级因果干预模块，通过原型字典主动消除混淆因素。
- 在nuScenes数据集上实现顶尖规划准确率和安全性表现。
- 对数据偏差和噪声场景具有强鲁棒性，适合高安全要求的自动驾驶应用。
以规划为导向的端到端驾驶模型展现出巨大潜力，但其本质学习的是统计相关性而非真正的因果关系。这种脆弱性导致因果混淆，使模型利用数据集偏差作为捷径，严重损害其在复杂场景中的可靠性与安全性。为此，我们提出CausalVAD，一种基于因果干预的去混淆训练框架。核心是设计稀疏因果干预方案（SCIS），一个轻量、即插即用的模块，用于在神经网络中实现后门调整理论。SCIS构建表示潜在驾驶情境的原型字典，并利用该字典对模型的稀疏向量查询进行干预，主动消除由混淆因子引发的虚假关联，从而从表征中移除虚假因素，用于下游任务。在nuScenes等基准上的大量实验表明，CausalVAD实现了最先进的规划准确率与安全性。此外，该方法在诱导因果混淆的数据偏差和噪声场景中表现出卓越鲁棒性。
原文摘要 · Abstract (English)
Planning-oriented end-to-end driving models show great promise, yet they fundamentally learn statistical correlations instead of true causal relationships. This vulnerability leads to causal confusion, where models exploit dataset biases as shortcuts, critically harming their reliability and safety in complex scenarios. To address this, we introduce CausalVAD, a de-confounding training framework that leverages causal intervention. At its core, we design the sparse causal intervention scheme (SCIS), a lightweight, plug-and-play module to instantiate the backdoor adjustment theory in neural networks. SCIS constructs a dictionary of prototypes representing latent driving contexts. It then uses this dictionary to intervene on the model's sparse vectorized queries. This step actively eliminates spurious associations induced by confounders, thereby eliminating spurious factors from the representations for downstream tasks. Extensive experiments on benchmarks like nuScenes show CausalVAD achieves state-of-the-art planning accuracy and safety. Furthermore, our method demonstrates superior robustness against both data bias and noisy scenarios configured to induce causal confusion.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。