Edge-Oriented Lightweight YOLO-World with Cross-Modal Fusion Adaptation for Multimodal Object Detection and Deployment
Applied Sciences · 05.08.2026 · 16 (15) · 7790
DOI: 10.3390/app16157790 ↗Kurz gefasst
Fachartikel in der Zeitschrift Applied Sciences, online erschienen am 05.08.2026. Verfasst von Mengnan Jiang und 3 weiteren Autor:innen. Die Arbeit nennt 36 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Multimodal detection models enable flexible object detection through text prompts, but YOLO-World-style models still incur high computational and storage costs on edge devices. To address this problem, this paper develops an edge-oriented adaptation framework for replacing the original visual encoder of YOLOv8l-worldv2 with a compact visual branch. The framework jointly considers visual-branch compression, multi-scale interface consistency, cross-modal feature compatibility, and edge-side inference, rather than optimizing these aspects independently. Specifically, a YOLOv7-based visual branch is reconstructed using depthwise separable convolutions, enhanced by DyHead, and compressed through sensitivity-guided Filter Pruning via Geometric Median (FPGM) under multi-scale interface constraints. An identity-initialized semantic adaptation layer and multi-template text prototypes are then introduced to alleviate the feature distribution mismatch between the compressed visual branch and the original cross-modal fusion space. Finally, offline text prototype generation and TensorRT-based INT8/FP16 mixed-precision inference are used for edge deployment. Experiments on BDD100K show that the lightweight visual encoder achieves 62.45% mAP@0.5 with 22.5 M parameters and 52.8 GFLOPs. After multimodal integration, the proposed model achieves 64.9% mAP@0.5 with 31.5 M parameters and 67.2 GFLOPs, reducing GFLOPs by 67.1% compared with YOLOv8l-worldv2 while causing only a 1.6 percentage-point accuracy drop. On the Jetson Orin Nano Super, the deployed model reaches 32.6 FPS with a model size of 32 MB, demonstrating its feasibility for edge perception scenarios.
Zitieren
APA
Jiang, M., Mo, T., Hu, J., & Liu, Y. (2026). Edge-Oriented Lightweight YOLO-World with Cross-Modal Fusion Adaptation for Multimodal Object Detection and Deployment. Applied Sciences, 16(15), 7790. https://doi.org/10.3390/app16157790
Vancouver
Jiang M, Mo T, Hu J, Liu Y. Edge-Oriented Lightweight YOLO-World with Cross-Modal Fusion Adaptation for Multimodal Object Detection and Deployment. Applied Sciences. 2026;16(15):7790. doi:10.3390/app16157790
BibTeX
@article{Jiang2026,
author = {Jiang, Mengnan and Mo, Tianli and Hu, Jie and Liu, Yanyi},
title = {Edge-Oriented Lightweight YOLO-World with Cross-Modal Fusion Adaptation for Multimodal Object Detection and Deployment},
journal = {Applied Sciences},
year = {2026},
volume = {16},
number = {15},
pages = {7790},
doi = {10.3390/app16157790}
}
EUVIMED