Vehicle re-identification based on dimensional decoupling strategy and non-local relations
PLOS ONE · 02.01.2024 · 19 (1) · e0291047
DOI: 10.1371/journal.pone.0291047 ↗Kurz gefasst
Fachartikel in der Zeitschrift PLOS ONE, online erschienen am 02.01.2024. Verfasst von Xi Li und 2 weiteren Autor:innen. Beteiligte Einrichtung: School of Information Science and Electrical Engineering, Shandong Jiaotong University, Jinan, Shan Dong, China. Hauptthemen laut MEDLINE: Endocrinologists, Lighting. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 56 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Focus on Research and Development Plan in Shandong Province.
Abstract (Originalsprache)
Vehicle re-identification (Re-ID) is a challenging task that aims to recognize the same vehicle across different non-overlapping cameras. Existing attention mechanism-based methods for vehicle Re-ID often suffer from significant intra-class variation and inter-class variation due to various factors such as illumination, occlusion, viewpoint, etc. In this paper, we propose a novel network architecture for vehicle Re-ID, named Dimensional Decoupling Strategy and Non-local Relationship Network (DMNR-Net), which uses three modules to extract complementary features: global feature extraction module, non-local relationship capture module(NRCM), and dimensional decoupling module (DDS). The global feature extraction module captures complete and coarse-grained features from the whole image; the NRCM module extracts saliency information from feature maps in both spatial and channel dimensions; and the DDS decouples spatial and channel features into two branches to extract fine-grained features and focus on specific subspaces. We conduct extensive experiments on two popular publicly datasets, VeRi-776 and VehicleID, to evaluate the effectiveness of our method. The experimental results show that our DMNR-Net outperforms state-of-the-art methods by a large margin on both datasets.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- School of Information Science and Electrical Engineering, Shandong Jiaotong University, Jinan, Shan Dong, China.
ORCID-Profile: Xiyu Pang
Förderung
- Focus on Research and Development Plan in Shandong Province
Zitieren
APA
Li, X., Pang, X., & Meng, Q. (2024). Vehicle re-identification based on dimensional decoupling strategy and non-local relations. PLOS ONE, 19(1), e0291047. https://doi.org/10.1371/journal.pone.0291047
Vancouver
Li X, Pang X, Meng Q. Vehicle re-identification based on dimensional decoupling strategy and non-local relations. PLOS ONE. 2024;19(1):e0291047. doi:10.1371/journal.pone.0291047
BibTeX
@article{Li2024,
author = {Li, Xi and Pang, Xiyu and Meng, Qinglan},
title = {Vehicle re-identification based on dimensional decoupling strategy and non-local relations},
journal = {PLOS ONE},
year = {2024},
volume = {19},
number = {1},
pages = {e0291047},
doi = {10.1371/journal.pone.0291047}
}
EUVIMED