SparScene: Efficient Traffic Scene Representation via Sparse Graph Learning for Large-Scale Trajectory Generation
IEEE Transactions on Pattern Analysis and Machine Intelligence · 2026 · 1-16
DOI: 10.1109/tpami.2026.3750173 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Pattern Analysis and Machine Intelligence, 2026. Verfasst von Xiaoyu Mo und 4 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Multi-agent trajectory generation is a core problem for autonomous driving and intelligent transportation systems. However, efficiently modeling the dynamic interactions between numerous road users and infrastructures in complex scenes remains an open problem. Existing methods typically employ distance-based or fully connected dense graph structures to capture interaction information, which not only introduces a large number of redundant edges but also requires complex and heavily parameterized networks for encoding, thereby resulting in low training and inference efficiency, limiting scalability to large and complex traffic scenes. To overcome the limitations of existing methods, we propose SparScene, a sparse graph learning framework designed for efficient and computationally scalable traffic scene representation. Instead of relying on distance thresholds, SparScene leverages the lane graph topology to construct structure-aware sparse connections between agents and lanes, enabling efficient yet informative scene graph representation. SparScene adopts a lightweight graph encoder that efficiently aggregates agent-map and agent-agent interactions, yielding compact scene representations with substantially improved efficiency and inference scalability. On the motion prediction benchmark of the Waymo Open Motion Dataset, SparScene achieves competitive performance with remarkable efficiency. It generates trajectories for more than 200 agents in a scene within 5 ms and scales to more than 5,000 agents and 17,000 lanes with merely 54 ms of inference time with only 2.9 GB of GPU memory, highlighting its superior computational scalability for large-scale traffic scenes.
Zitieren
APA
Mo, X., Ge, J., Wang, Z., Lv, C., & Johansson, K. H. (2026). SparScene: Efficient Traffic Scene Representation via Sparse Graph Learning for Large-Scale Trajectory Generation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1-16. https://doi.org/10.1109/tpami.2026.3750173
Vancouver
Mo X, Ge J, Wang Z, Lv C, Johansson KH. SparScene: Efficient Traffic Scene Representation via Sparse Graph Learning for Large-Scale Trajectory Generation. IEEE Transactions on Pattern Analysis and Machine Intelligence. 2026:1-16. doi:10.1109/tpami.2026.3750173
BibTeX
@article{Mo2026,
author = {Mo, Xiaoyu and Ge, Jintian and Wang, Zifan and Lv, Chen and Johansson, Karl H},
title = {SparScene: Efficient Traffic Scene Representation via Sparse Graph Learning for Large-Scale Trajectory Generation},
journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence},
year = {2026},
pages = {1--16},
doi = {10.1109/tpami.2026.3750173}
}
EUVIMED