通过显式建模帧率信息，实现跨场景零样本轨迹预测
OmniTraj: Pre-Training on Heterogeneous Data for Adaptive and Zero-Shot Human Trajectory Prediction
- 用帧率等时间元数据显式条件化，提升模型泛化能力
- 零样本迁移下预测误差降低70%以上，优于现有方法
- 适合需要快速部署到新数据集的智能交通与机器人应用
尽管大规模预训练推动了人类轨迹预测发展，但一个关键挑战仍存：在不同时间动态的新数据集上实现零样本迁移。当前先进模型常需微调才能适应不同帧率或观测时长的数据集，限制了其可扩展性与实用性。本文系统研究此问题，发现现有数据感知离散模型在时间设置变化时表现不佳。通过分离时间泛化与数据集偏移，揭示显式时间元数据条件化是有效解法。基于此，提出OmniTraj——一个在大规模异构数据上预训练的Transformer模型。实验表明，显式依赖帧率使OmniTraj在复杂跨设置场景中零样本性能达最优，预测误差下降超70%。微调后在NBA、JTA、WorldPose和ETH-UCY四个数据集上均达到领先水平。代码已公开：https://github.com/vita-epfl/omnitraj
原文摘要 · Abstract (English)
While large-scale pre-training has advanced human trajectory prediction, a critical challenge remains: zero-shot transfer to unseen dataset with varying temporal dynamics. State-of-the-art pre-trained models often require fine-tuning to adapt to new datasets with different frame rates or observation horizons, limiting their scalability and practical utility. In this work, we systematically investigate this limitation and propose a robust solution. We first demonstrate that existing data-aware discrete models struggle when transferred to new scenarios with shifted temporal setups. We then isolate the temporal generalization from dataset shift, revealing that a simple, explicit conditioning mechanism for temporal metadata is a highly effective solution. Based on this insight, we present OmniTraj, a Transformer-based model pre-trained on a large-scale, heterogeneous dataset. Our experiments show that explicitly conditioning on the frame rate enables OmniTraj to achieve state-of-the-art zero-shot transfer performance, reducing prediction error by over 70\% in challenging cross-setup scenarios. After fine-tuning, OmniTraj achieves state-of-the-art results on four datasets, including NBA, JTA, WorldPose, and ETH-UCY. The code is publicly available: https://github.com/vita-epfl/omnitraj
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。