OmniTraj: Pre-Training on Heterogeneous Data for Adaptive and Zero-Shot Human Trajectory Prediction

TL;DR

OmniTraj is a Transformer-based human trajectory prediction model pre-trained on diverse data, which effectively generalizes to unseen datasets with different temporal dynamics without fine-tuning, significantly improving zero-shot transfer performance.

Contribution

The paper introduces OmniTraj, a novel pre-trained model that explicitly conditions on temporal metadata, enabling robust zero-shot transfer across datasets with varying temporal setups.

Findings

Achieves over 70% reduction in prediction error in cross-setup scenarios.

Outperforms existing models on four benchmark datasets after fine-tuning.

Explicit temporal conditioning is key to generalization across different temporal dynamics.

Abstract

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