CAFE-AD: Cross-Scenario Adaptive Feature Enhancement for Trajectory Planning in Autonomous Driving

TL;DR

CAFE-AD tackles the problem of causal confusion and overfitting in imitation-learning-based trajectory planning for autonomous driving by introducing Cross-Scenario Adaptive Feature Enhancement. It comprises an adaptive feature pruning module to isolate the most informative features and a cross-scenario feature interpolation module to diversify scenario representations and combat long-tail biases. Evaluations on the nuPlan Test14-Hard closed-loop simulation benchmark show that CAFE-AD outperforms state-of-the-art rule-based and hybrid planners, and mitigates the impact of long-tail distributions; real-world experiments further validate the approach. The work provides code and models, underscoring practical applicability and potential to improve reliable planning across diverse driving scenarios.

Abstract

Imitation learning based planning tasks on the nuPlan dataset have gained great interest due to their potential to generate human-like driving behaviors. However, open-loop training on the nuPlan dataset tends to cause causal confusion during closed-loop testing, and the dataset also presents a long-tail distribution of scenarios. These issues introduce challenges for imitation learning. To tackle these problems, we introduce CAFE-AD, a Cross-Scenario Adaptive Feature Enhancement for Trajectory Planning in Autonomous Driving method, designed to enhance feature representation across various scenario types. We develop an adaptive feature pruning module that ranks feature importance to capture the most relevant information while reducing the interference of noisy information during training. Moreover, we propose a cross-scenario feature interpolation module that enhances scenario information to introduce diversity, enabling the network to alleviate over-fitting in dominant scenarios. We evaluate our method CAFE-AD on the challenging public nuPlan Test14-Hard closed-loop simulation benchmark. The results demonstrate that CAFE-AD outperforms state-of-the-art methods including rule-based and hybrid planners, and exhibits the potential in mitigating the impact of long-tail distribution within the dataset. Additionally, we further validate its effectiveness in real-world environments. The code and models will be made available at https://github.com/AlniyatRui/CAFE-AD.