Global Selection and Local Differentiation Fusion for Vehicle Re-identification

Authors

Shengjun Xu, College of Information and Control Engineering, Xi 'an University of Architecture and Technology, Xi'an 710055, China; Xi'an Key Labratory of Building Manufactaring Intelligent & Automation Technology, Xi'an 710055, China
Mengqian Zhang, College of Information and Control Engineering, Xi 'an University of Architecture and Technology, Xi'an 710055, China; Xi'an Key Labratory of Building Manufactaring Intelligent & Automation Technology, Xi'an 710055, China
Bohan Zhan, Xi'an Jiaotong University, Xi'an 710049, China
Guanghui Liu, College of Information and Control Engineering, Xi 'an University of Architecture and Technology, Xi'an 710055, China; Xi'an Key Labratory of Building Manufactaring Intelligent & Automation Technology, Xi'an 710055, China
Yuebo Meng, College of Information and Control Engineering, Xi 'an University of Architecture and Technology, Xi'an 710055, China; Xi'an Key Labratory of Building Manufactaring Intelligent & Automation Technology, Xi'an 710055, China

Abstract

Abstract: To address the interference issues of different perspectives, complex backgrounds, and lighting intensity in vehicle re-identification caused by cross lens multi view differences, a vehicle reidentification network integrating global selection and local differentiation is proposed. Based on Resnet50 backbone network, a three-branch complementary network integrating global and local features is designed. The global branch is used to learn overall appearance information of the vehicle, while the local branch captures differential details of the vehicle. Based on attention mechanism, a context feature selection module (CFSM) is proposed to effectively separate vehicle information from complex background information, and a detail feature enhancement module (DFEM) is proposed to enhance the learning of multi granularity feature detail information by utilizing relative position information between components. A weight adaptive balancing strategy is proposed in combination with multiple loss functions for training. Experimental results show that on VeRi-776 dataset, the proposed network achieves 73.2%, 93.4% and 97.3% respectively on the mAP, CMC@1 and CMC@5. On a largescale testing subset of VehicleID dataset, the proposed network achieves 75.0% and 92.7% respectively on the CMC@1 and CMC@5. Compared with the comparison network, the proposed network has higher recognition rate and robustness.

Recommended Citation

Xu, Shengjun; Zhang, Mengqian; Zhan, Bohan; Liu, Guanghui; and Meng, Yuebo (2025) "Global Selection and Local Differentiation Fusion for Vehicle Re-identification," Journal of System Simulation: Vol. 37: Iss. 1, Article 18.
DOI: 10.16182/j.issn1004731x.joss.23-1009
Available at: https://dc-china-simulation.researchcommons.org/journal/vol37/iss1/18

First Page

220

Last Page

233

CLC

TP391.9; TP181

Recommended Citation

Xu Shengjun, Zhang Mengqian, Zhan Bohan, et al. Global Selection and Local Differentiation Fusion for Vehicle Re-identification[J]. Journal of System Simulation, 2025, 37(1): 220-233.

Corresponding Author

Zhang Mengqian

DOI

10.16182/j.issn1004731x.joss.23-1009