SAIL: Scene-aware adaptive iterative learning for long-tail trajectory prediction in autonomous vehicles
Transportation Research Part C: Emerging Technologies · 2026-09 · 190 · 105757
DOI: 10.1016/j.trc.2026.105757 ↗Kurz gefasst
Fachartikel in der Zeitschrift Transportation Research Part C: Emerging Technologies, 2026-09. Verfasst von Bin Rao und 5 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 73 Literaturangaben. Gefördert von: University of Macau Research Services and Knowledge Transfer Office, Science and Technology Planning Project of Guangdong Province, Fundo para o Desenvolvimento das Ciências e da Tecnologia.
Abstract (Originalsprache)
Autonomous vehicles (AVs) rely on accurate trajectory prediction for safe navigation in diverse traffic environments, yet existing models struggle with long-tail scenarios-rare but safety-critical events characterized by abrupt maneuvers, high collision risks, and complex interactions. These challenges stem from data imbalance, inadequate definitions of long-tail trajectories, and suboptimal learning strategies that prioritize common behaviors over infrequent ones. To address this, we propose SAIL, a novel framework that systematically tackles the long-tail problem by first defining and modeling trajectories across three key attribute dimensions: prediction error, collision risk, and state complexity. Our approach then synergizes an attribute-guided augmentation and feature extraction process with a highly adaptive contrastive learning strategy. This strategy employs a continuous cosine momentum schedule, similarity-weighted hard-negative mining, and a dynamic pseudo-labeling mechanism based on evolving feature clustering. Furthermore, it incorporates a focusing mechanism to intensify learning on hard-positive samples within each identified class. This comprehensive design enables SAIL to excel at identifying and forecasting diverse and challenging long-tail events. Extensive evaluations on the nuScenes and ETH/UCY datasets demonstrate SAIL's superior performance, achieving up to 28.8% reduction in prediction error on the hardest 1% of long-tail samples compared to state-of-the-art baselines, while maintaining competitive accuracy across all scenarios. This framework advances reliable AV trajectory prediction in real-world, mixed-autonomy settings.
Förderung
- University of Macau Research Services and Knowledge Transfer Office
- Science and Technology Planning Project of Guangdong Province
- Fundo para o Desenvolvimento das Ciências e da Tecnologia
- Universidade de Macau
- Shenzhen Science and Technology Innovation Program
- National Natural Science Foundation of China
Zitieren
APA
Rao, B., Liao, H., Wang, C., Li, K., Li, Z., & Yang, H. (2026). SAIL: Scene-aware adaptive iterative learning for long-tail trajectory prediction in autonomous vehicles. Transportation Research Part C: Emerging Technologies, 190, 105757. https://doi.org/10.1016/j.trc.2026.105757
Vancouver
Rao B, Liao H, Wang C, Li K, Li Z, Yang H. SAIL: Scene-aware adaptive iterative learning for long-tail trajectory prediction in autonomous vehicles. Transportation Research Part C: Emerging Technologies. 2026;190:105757. doi:10.1016/j.trc.2026.105757
BibTeX
@article{Rao2026,
author = {Rao, Bin and Liao, Haicheng and Wang, Chengyue and Li, Keqiang and Li, Zhenning and Yang, Hai},
title = {SAIL: Scene-aware adaptive iterative learning for long-tail trajectory prediction in autonomous vehicles},
journal = {Transportation Research Part C: Emerging Technologies},
year = {2026},
volume = {190},
pages = {105757},
doi = {10.1016/j.trc.2026.105757}
}
EUVIMED