先恢复再预测：面向变长轨迹预测的渐进式回溯学习
Recover to Predict: Progressive Retrospective Learning for Variable-Length Trajectory Prediction
针对变长/不完整观测下的轨迹预测，用级联回溯单元逐步把不完整观测特征对齐到完整观测特征，并以滚动起点训练提升数据效率；该「由短到长渐进恢复」思路可迁移到任何存在观测缺失的时序视觉预测任务。
原文依据 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
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论轨迹预测对自动驾驶至关重要，它使车辆能够在密集、动态的交通环境中进行安全高效的规划。现有方法大多在固定长度观测的条件下优化预测精度。然而，真实驾驶场景往往产生变长、不完整的观测，这给上述方法带来了挑战。一种常见策略是将不完整观测的特征直接映射为完整观测的特征。然而，由于信息差距显著，这种单样本映射难以针对短轨迹学习到准确的表征。为解决该问题，我们提出渐进式回溯框架（PRF），它通过级联的回溯单元，将不完整观测的特征逐步与完整观测的特征对齐。每个单元由回溯蒸馏模块（RDM）与回溯预测模块（RPM）组成，其中 RDM 负责蒸馏特征，RPM 则利用蒸馏后的特征恢复此前的时刻。此外，我们提出滚动起点训练策略（RSTS），以提升 PRF 训练过程中的数据效率。PRF 可与现有方法即插即用。在 Argoverse 2 与 Argoverse 1 数据集上的大量实验验证了 PRF 的有效性。代码已发布于 https://github.com/zhouhao94/PRF。
英文原文摘要 1288 字符
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.