PlanScope: Learning to Plan Within Decision Scope for Urban Autonomous Driving

TL;DR

PlanScope introduces a novel framework for urban autonomous driving that separates short-term and long-term decision-making, improving planning accuracy by leveraging wavelet transforms and multi-scope supervision.

Contribution

The paper presents a new decision separation method using wavelet transforms and multi-scope supervision to enhance imitation learning for autonomous driving.

Findings

Outperforms baseline models in closed-loop evaluations

Effective separation of decision components improves planning accuracy

Time-dependent normalization enhances neural network detail generation

Abstract

In the context of urban autonomous driving, imitation learning-based methods have shown remarkable effectiveness, with a typical practice to minimize the discrepancy between expert driving logs and predictive decision sequences. As expert driving logs natively contain future short-term decisions with respect to events, such as sudden obstacles or rapidly changing traffic signals. We believe that unpredictable future events and corresponding expert reactions can introduce reasoning disturbances, negatively affecting the convergence efficiency of planning models. At the same time, long-term decision information, such as maintaining a reference lane or avoiding stationary obstacles, is essential for guiding short-term decisions. Our preliminary experiments on shortening the planning horizon show a rise-and-fall trend in driving performance, supporting these hypotheses. Based on these insights, we present PlanScope, a sequential-decision-learning framework with novel techniques for separating short-term and long-term decisions in decision logs. To identify and extract each decision component, the Wavelet Transform on trajectory profiles is proposed. After that, to enhance the detail-generating ability of Neural Networks, extra Detail Decoders are proposed. Finally, to enable in-scope decision supervision across detail levels, Multi-Scope Supervision strategies are adopted during training. The proposed methods, especially the time-dependent normalization, outperform baseline models in closed-loop evaluations on the nuPlan dataset, offering a plug-and-play solution to enhance existing planning models.