ViTE: Virtual Graph Trajectory Expert Router for Pedestrian Trajectory Prediction

Ruochen Li; Zhanxing Zhu; Tanqiu Qiao; Hubert P. H. Shum

AAAI 2026technical0 citations

ViTE: Virtual Graph Trajectory Expert Router for Pedestrian Trajectory Prediction

Ruochen Li, Zhanxing Zhu, Tanqiu Qiao, Hubert P. H. Shum

Abstract

Pedestrian trajectory prediction is critical for ensuring safety in autonomous driving, surveillance systems, and urban planning applications. While early approaches primarily focus on one-hop pairwise relationships, recent studies attempt to capture high-order interactions by stacking multiple Graph Neural Network (GNN) layers. However, these approaches face a fundamental trade-off: insufficient layers may lead to under-reaching problems that limit the model

BibTeX

@inproceedings{aaai2026_vitevirtualgraph,
  title = {ViTE: Virtual Graph Trajectory Expert Router for Pedestrian Trajectory Prediction},
  author = {Ruochen Li and Zhanxing Zhu and Tanqiu Qiao and Hubert P. H. Shum},
  booktitle = {AAAI 2026},
  year = {2026}
}