HSF-DETR: A Special Vehicle Detection Algorithm Based on Hypergraph Spatial Features and Bipolar Attention

Special vehicle detection in intelligent surveillance, emergency rescue, and reconnaissance faces significant challenges in accuracy and robustness under complex environments, necessitating advanced detection algorithms for critical applications. This paper proposes HSF-DETR (Hypergraph Spatial Feat...

Full description

Saved in:
Bibliographic Details
Main Authors: Kaipeng Wang, Guanglin He, Xinmin Li
Format: Article
Language:English
Published: MDPI AG 2025-07-01
Series:Sensors
Subjects:
Online Access:https://www.mdpi.com/1424-8220/25/14/4381
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849251766994993152
author Kaipeng Wang
Guanglin He
Xinmin Li
author_facet Kaipeng Wang
Guanglin He
Xinmin Li
author_sort Kaipeng Wang
collection DOAJ
description Special vehicle detection in intelligent surveillance, emergency rescue, and reconnaissance faces significant challenges in accuracy and robustness under complex environments, necessitating advanced detection algorithms for critical applications. This paper proposes HSF-DETR (Hypergraph Spatial Feature DETR), integrating four innovative modules: a Cascaded Spatial Feature Network (CSFNet) backbone with Cross-Efficient Convolutional Gating (CECG) for enhanced long-range detection through hybrid state-space modeling; a Hypergraph-Enhanced Spatial Feature Modulation (HyperSFM) network utilizing hypergraph structures for high-order feature correlations and adaptive multi-scale fusion; a Dual-Domain Feature Encoder (DDFE) combining Bipolar Efficient Attention (BEA) and Frequency-Enhanced Feed-Forward Network (FEFFN) for precise feature weight allocation; and a Spatial-Channel Fusion Upsampling Block (SCFUB) improving feature fidelity through depth-wise separable convolution and channel shift mixing. Experiments conducted on a self-built special vehicle dataset containing 2388 images demonstrate that HSF-DETR achieves mAP50 and mAP50-95 of 96.6% and 70.6%, respectively, representing improvements of 3.1% and 4.6% over baseline RT-DETR while maintaining computational efficiency at 59.7 GFLOPs and 18.07 M parameters. Cross-domain validation on VisDrone2019 and BDD100K datasets confirms the method’s generalization capability and robustness across diverse scenarios, establishing HSF-DETR as an effective solution for special vehicle detection in complex environments.
format Article
id doaj-art-56e20376d9af4390aaeddfbc3a50ecdd
institution Kabale University
issn 1424-8220
language English
publishDate 2025-07-01
publisher MDPI AG
record_format Article
series Sensors
spelling doaj-art-56e20376d9af4390aaeddfbc3a50ecdd2025-08-20T03:56:49ZengMDPI AGSensors1424-82202025-07-012514438110.3390/s25144381HSF-DETR: A Special Vehicle Detection Algorithm Based on Hypergraph Spatial Features and Bipolar AttentionKaipeng Wang0Guanglin He1Xinmin Li2Science and Technology on Electromechanical Dynamic Control Laboratory, Beijing Institute of Technology, Beijing 100081, ChinaScience and Technology on Electromechanical Dynamic Control Laboratory, Beijing Institute of Technology, Beijing 100081, ChinaScience and Technology on Electromechanical Dynamic Control Laboratory, Beijing Institute of Technology, Beijing 100081, ChinaSpecial vehicle detection in intelligent surveillance, emergency rescue, and reconnaissance faces significant challenges in accuracy and robustness under complex environments, necessitating advanced detection algorithms for critical applications. This paper proposes HSF-DETR (Hypergraph Spatial Feature DETR), integrating four innovative modules: a Cascaded Spatial Feature Network (CSFNet) backbone with Cross-Efficient Convolutional Gating (CECG) for enhanced long-range detection through hybrid state-space modeling; a Hypergraph-Enhanced Spatial Feature Modulation (HyperSFM) network utilizing hypergraph structures for high-order feature correlations and adaptive multi-scale fusion; a Dual-Domain Feature Encoder (DDFE) combining Bipolar Efficient Attention (BEA) and Frequency-Enhanced Feed-Forward Network (FEFFN) for precise feature weight allocation; and a Spatial-Channel Fusion Upsampling Block (SCFUB) improving feature fidelity through depth-wise separable convolution and channel shift mixing. Experiments conducted on a self-built special vehicle dataset containing 2388 images demonstrate that HSF-DETR achieves mAP50 and mAP50-95 of 96.6% and 70.6%, respectively, representing improvements of 3.1% and 4.6% over baseline RT-DETR while maintaining computational efficiency at 59.7 GFLOPs and 18.07 M parameters. Cross-domain validation on VisDrone2019 and BDD100K datasets confirms the method’s generalization capability and robustness across diverse scenarios, establishing HSF-DETR as an effective solution for special vehicle detection in complex environments.https://www.mdpi.com/1424-8220/25/14/4381special vehicle detectionRT-DETRmulti-scale feature fusiondeep learningobject detection
spellingShingle Kaipeng Wang
Guanglin He
Xinmin Li
HSF-DETR: A Special Vehicle Detection Algorithm Based on Hypergraph Spatial Features and Bipolar Attention
Sensors
special vehicle detection
RT-DETR
multi-scale feature fusion
deep learning
object detection
title HSF-DETR: A Special Vehicle Detection Algorithm Based on Hypergraph Spatial Features and Bipolar Attention
title_full HSF-DETR: A Special Vehicle Detection Algorithm Based on Hypergraph Spatial Features and Bipolar Attention
title_fullStr HSF-DETR: A Special Vehicle Detection Algorithm Based on Hypergraph Spatial Features and Bipolar Attention
title_full_unstemmed HSF-DETR: A Special Vehicle Detection Algorithm Based on Hypergraph Spatial Features and Bipolar Attention
title_short HSF-DETR: A Special Vehicle Detection Algorithm Based on Hypergraph Spatial Features and Bipolar Attention
title_sort hsf detr a special vehicle detection algorithm based on hypergraph spatial features and bipolar attention
topic special vehicle detection
RT-DETR
multi-scale feature fusion
deep learning
object detection
url https://www.mdpi.com/1424-8220/25/14/4381
work_keys_str_mv AT kaipengwang hsfdetraspecialvehicledetectionalgorithmbasedonhypergraphspatialfeaturesandbipolarattention
AT guanglinhe hsfdetraspecialvehicledetectionalgorithmbasedonhypergraphspatialfeaturesandbipolarattention
AT xinminli hsfdetraspecialvehicledetectionalgorithmbasedonhypergraphspatialfeaturesandbipolarattention