MetaTra: Meta-Learning for Generalized Trajectory Prediction in Unseen Domain
This addresses the challenge of generalizing trajectory prediction for autonomous driving and robotics in new environments without model updates, representing a novel method for a known bottleneck.
The paper tackles the problem of trajectory prediction models failing in unseen domains by proposing MetaTra, a meta-learning method that achieves state-of-the-art performance on real-world datasets with plug-and-play capabilities for domain generalization.
Trajectory prediction has garnered widespread attention in different fields, such as autonomous driving and robotic navigation. However, due to the significant variations in trajectory patterns across different scenarios, models trained in known environments often falter in unseen ones. To learn a generalized model that can directly handle unseen domains without requiring any model updating, we propose a novel meta-learning-based trajectory prediction method called MetaTra. This approach incorporates a Dual Trajectory Transformer (Dual-TT), which enables a thorough exploration of the individual intention and the interactions within group motion patterns in diverse scenarios. Building on this, we propose a meta-learning framework to simulate the generalization process between source and target domains. Furthermore, to enhance the stability of our prediction outcomes, we propose a Serial and Parallel Training (SPT) strategy along with a feature augmentation method named MetaMix. Experimental results on several real-world datasets confirm that MetaTra not only surpasses other state-of-the-art methods but also exhibits plug-and-play capabilities, particularly in the realm of domain generalization.