Adapting to Length Shift: FlexiLength Network for Trajectory Prediction

Yi Xu, Yun Fu

Abstract: Trajectory prediction plays an important role in various applications, including autonomous driving, robotics, and scene understanding. Existing approaches mainly focus on developing compact neural networks to increase prediction precision on public datasets, typically employing a stan-dardized input duration. However, a notable issue arises when these models are evaluated with varying observation lengths, leading to a significant performance drop, a phe-nomenon we term the Observation Length Shift. To address this issue, we introduce a general and effective framework, the FlexiLength Network (FLN), to enhance the robustness of existing trajectory prediction techniques against varying observation periods. Specifically, FLN integrates tra-jectory data with diverse observation lengths, incorporates FlexiLength Calibration (FLC) to acquire temporal invari-ant representations, and employs FlexiLength Adaptation (FLA) to further refine these representations for more ac-curate future trajectory predictions. Comprehensive exper-iments on multiple datasets, i.e., ETH/UCY, nuScenes, and Argoverse 1, demonstrate the effectiveness and flexibility of our proposed FLN framework.