Recover to Predict: Progressive Retrospective Learning for Variable-Length Trajectory Prediction

TL;DR

This paper introduces a Progressive Retrospective Framework (PRF) that improves variable-length trajectory prediction by gradually aligning incomplete and complete observations through a cascade of modules, enhancing accuracy in autonomous driving scenarios.

Contribution

The paper proposes a novel PRF with retrospective units and a rolling-start training strategy, enabling better feature learning from incomplete observations in trajectory prediction.

Findings

PRF significantly improves prediction accuracy on Argoverse datasets.

The method effectively handles variable-length, incomplete trajectory data.

PRF is compatible with existing prediction models.

Abstract

Trajectory prediction is critical for autonomous driving, enabling safe and efficient planning in dense, dynamic traffic. Most existing methods optimize prediction accuracy under fixed-length observations. However, real-world driving often yields variable-length, incomplete observations, posing a challenge to these methods. A common strategy is to directly map features from incomplete observations to those from complete ones. This one-shot mapping, however, struggles to learn accurate representations for short trajectories due to significant information gaps. To address this issue, we propose a Progressive Retrospective Framework (PRF), which gradually aligns features from incomplete observations with those from complete ones via a cascade of retrospective units. Each unit consists of a Retrospective Distillation Module (RDM) and a Retrospective Prediction Module (RPM), where RDM distills features and RPM recovers previous timesteps using the distilled features. Moreover, we propose a Rolling-Start Training Strategy (RSTS) that enhances data efficiency during PRF training. PRF is plug-and-play with existing methods. Extensive experiments on datasets Argoverse 2 and Argoverse 1 demonstrate the effectiveness of PRF. Code is available at https://github.com/zhouhao94/PRF.