DTPP: Differentiable Joint Conditional Prediction and Cost Evaluation for Tree Policy Planning in Autonomous Driving

TL;DR

This paper introduces a differentiable joint training framework for motion prediction and cost evaluation in autonomous driving, improving planning performance and efficiency through a tree-structured policy and novel models.

Contribution

It proposes a novel differentiable joint training approach for prediction and cost models, utilizing a query-centric Transformer and a context-aware cost function for autonomous vehicle planning.

Findings

Outperforms state-of-the-art planning methods in quality

Operates more efficiently in runtime

Joint training significantly improves performance

Abstract

Motion prediction and cost evaluation are vital components in the decision-making system of autonomous vehicles. However, existing methods often ignore the importance of cost learning and treat them as separate modules. In this study, we employ a tree-structured policy planner and propose a differentiable joint training framework for both ego-conditioned prediction and cost models, resulting in a direct improvement of the final planning performance. For conditional prediction, we introduce a query-centric Transformer model that performs efficient ego-conditioned motion prediction. For planning cost, we propose a learnable context-aware cost function with latent interaction features, facilitating differentiable joint learning. We validate our proposed approach using the real-world nuPlan dataset and its associated planning test platform. Our framework not only matches state-of-the-art planning methods but outperforms other learning-based methods in planning quality, while operating more efficiently in terms of runtime. We show that joint training delivers significantly better performance than separate training of the two modules. Additionally, we find that tree-structured policy planning outperforms the conventional single-stage planning approach.