Roadside Perception Fusion Algorithm Based on Transformer and Design of Lightweight Edge Computing Platform
Journal of Intelligence and Knowledge Engineering · 2026-06 · 4 (2) · 28-36
DOI: 10.62517/jike.202604204 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Intelligence and Knowledge Engineering, 2026-06. Verfasst von Liangdong Zuo und 4 weiteren Autor:innen. Die Arbeit nennt 15 Literaturangaben.
Abstract (Originalsprache)
To address challenges of roadside perception systems (single-sensor limitations, ineffective multi-modal fusion, incompatible heavy models with edge devices), this paper proposes a Transformer-based fusion algorithm and a lightweight edge platform. It designs a Multi-Modal Transformer Fusion (MMTF) network with dual encoders and cross-attention for adaptive feature fusion, lightweightens the network via knowledge distillation and layer pruning, and builds an edge platform based on NVIDIA Jetson Orin NX. Experiments on DAIR-V2X and TUMTraf Intersection datasets show the MMTF achieves 92.3% 3D object detection mAP (6.8%/9.2% higher than CNN-based/single-sensor algorithms), with the lightweight model reducing parameters by 65% and latency by 58%, and the platform running stably at 30 FPS (12W). This provides a high-performance, low-cost solution for roadside perception in ITS with significant theoretical and engineering value.
Zitieren
APA
Zuo, L., Li, J., Liu, J., Li, H., & Huang, M. (2026). Roadside Perception Fusion Algorithm Based on Transformer and Design of Lightweight Edge Computing Platform. Journal of Intelligence and Knowledge Engineering, 4(2), 28-36. https://doi.org/10.62517/jike.202604204
Vancouver
Zuo L, Li J, Liu J, Li H, Huang M. Roadside Perception Fusion Algorithm Based on Transformer and Design of Lightweight Edge Computing Platform. Journal of Intelligence and Knowledge Engineering. 2026;4(2):28-36. doi:10.62517/jike.202604204
BibTeX
@article{Zuo2026,
author = {Zuo, Liangdong and Li, Jie and Liu, Jia and Li, Hejia and Huang, Mingfei},
title = {Roadside Perception Fusion Algorithm Based on Transformer and Design of Lightweight Edge Computing Platform},
journal = {Journal of Intelligence and Knowledge Engineering},
year = {2026},
volume = {4},
number = {2},
pages = {28--36},
doi = {10.62517/jike.202604204}
}
EUVIMED