ART: Adaptive Relational Transformer for Pedestrian Trajectory Prediction with Temporal-Aware Relations
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915915817287680 |
|---|---|
| author | Li, Ruochen Chang, Ziyi Hu, Junyan Li, Jiannan Atapour-Abarghouei, Amir Shum, Hubert P. H. |
| author_facet | Li, Ruochen Chang, Ziyi Hu, Junyan Li, Jiannan Atapour-Abarghouei, Amir Shum, Hubert P. H. |
| contents | Accurate prediction of real-world pedestrian trajectories is crucial for a wide range of robot-related applications. Recent approaches typically adopt graph-based or transformer-based frameworks to model interactions. Despite their effectiveness, these methods either introduce unnecessary computational overhead or struggle to represent the diverse and time-varying characteristics of human interactions. In this work, we present an Adaptive Relational Transformer (ART), which introduces a Temporal-Aware Relation Graph (TARG) to explicitly capture the evolution of pairwise interactions and an Adaptive Interaction Pruning (AIP) mechanism to reduce redundant computations efficiently. Extensive evaluations on ETH/UCY and NBA benchmarks show that ART delivers state-of-the-art accuracy with high computational efficiency. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_03649 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | ART: Adaptive Relational Transformer for Pedestrian Trajectory Prediction with Temporal-Aware Relations Li, Ruochen Chang, Ziyi Hu, Junyan Li, Jiannan Atapour-Abarghouei, Amir Shum, Hubert P. H. Computer Vision and Pattern Recognition Artificial Intelligence Accurate prediction of real-world pedestrian trajectories is crucial for a wide range of robot-related applications. Recent approaches typically adopt graph-based or transformer-based frameworks to model interactions. Despite their effectiveness, these methods either introduce unnecessary computational overhead or struggle to represent the diverse and time-varying characteristics of human interactions. In this work, we present an Adaptive Relational Transformer (ART), which introduces a Temporal-Aware Relation Graph (TARG) to explicitly capture the evolution of pairwise interactions and an Adaptive Interaction Pruning (AIP) mechanism to reduce redundant computations efficiently. Extensive evaluations on ETH/UCY and NBA benchmarks show that ART delivers state-of-the-art accuracy with high computational efficiency. |
| title | ART: Adaptive Relational Transformer for Pedestrian Trajectory Prediction with Temporal-Aware Relations |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2604.03649 |