

浏览全部资源
扫码关注微信
1.无锡学院 交通与车辆工程学院, 江苏 无锡 214105
2.南京信息工程大学 计算机学院, 江苏 南京 210044
3.无锡学院 物联网工程学院, 江苏 无锡 214105
4.江南大学 光电信息与物理科学学院, 江苏 无锡 214122
Received:29 June 2026,
Revised:2026-07-17,
Published:05 August 2026
移动端阅览
XU Mengxi, CHEN Haixin, XU Huanyu, et al. Vehicle re-identification network based on adaptive side information and residual hierarchical attention[J]. Chinese Journal of Liquid Crystals and Displays, 2026, 41(8): 1200-1212.
XU Mengxi, CHEN Haixin, XU Huanyu, et al. Vehicle re-identification network based on adaptive side information and residual hierarchical attention[J]. Chinese Journal of Liquid Crystals and Displays, 2026, 41(8): 1200-1212. DOI: 10.3724/CJLCD.2026-0118. CSTR: 32172.14.CJLCD.2026-0118.
车辆重识别旨在跨摄像头场景下准确匹配同一车辆,是智能交通与公共安全监控的重要任务。针对现有方法局部细粒度特征利用不足、侧信息融合静态以及多层注意力融合不稳定等问题,提出一种基于自适应侧信息与残差层级注意力的车辆重识别网络。该网络以Vision Transformer为基础架构,引入自适应侧信息整合模块,对Patch token进行动态调制与上下文增强,以缓解跨摄像头、跨视角条件下的特征分布偏移;设计稳定残差跨层注意力聚合模块,在多层可学习注意力融合中引入残差结构,提高多尺度语义融合稳定性和全局引导可靠性;构建多头层级自适应划分模块,自适应提取具有一定物理可解释性的局部表征。实验结果表明,该方法在VeRi-776数据集上取得83.8%的mAP、97.6%的Rank-1和98.8%的Rank-5;在VehicleID数据集Test-1600和Test-2400子集上分别取得80.6%和78.6%的Rank-1。本文方法在复杂车辆重识别场景中的准确性上具有一定优势,在大规模智能交通系统中具有潜在应用价值。
Vehicle re-identification aims to match the same vehicle across different camera views and is a critical task for intelligent transportation and public security surveillance. Aiming at the drawbacks of existing Vision Transformer-based vehicle re-identification methods, including insufficient mining of fine-grained local features, static side-information fusion strategies, and poor stability during multi-layer attention aggregation, this paper proposes a Vision Transformer vehicle re-identification network with adaptive side-information and residual hierarchical attention. Built upon the Vision Transformer backbone, the network embeds an Adaptive Side-information Integration module to dynamically modulate patch tokens and strengthen contextual cues, mitigating feature distribution discrepancies caused by varying cameras and viewing angles. Meanwhile, a Stable Residual Cross-layer Attention Aggregation module is devised, which embeds residual connections into multi-layer learnable attention fusion to stabilize multi-scale semantic aggregation and strengthen global feature guidance. In addition, a Multi-Head Hierarchical Adaptive Partitioning module is constructed to automatically extract local feature representations with a certain degree of spatial interpretability. Experimental results demonstrate that the proposed method reaches 83.8% mAP, 97.6% Rank-1 and 98.8% Rank-5 on the VeRi-776 dataset, and achieves Rank-1 accuracies of 80.6% and 78.6% on the Test-1600 and Test-2400 subsets of VehicleID respectively. The presented Vision Transformer model achieves remarkable accuracy in complicated vehicle re-identification scenarios, and possesses great practical application potential for large-scale intelligent transportation systems.
LIU X C , LIU W , MA H D , et al . Large-scale vehicle re-identification in urban surveillance videos [C]// Proceedings of the 2016 IEEE International Conference on Multimedia and Expo (ICME) . Seattle : IEEE , 2016 : 1 - 6 . doi: 10.1109/icme.2016.7553002 http://dx.doi.org/10.1109/icme.2016.7553002
LIU H Y , TIAN Y H , WANG Y W , et al . Deep relative distance learning: tell the difference between similar vehicles [C]// Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition . Las Vegas : IEEE , 2016 : 2167 - 2175 . doi: 10.1109/cvpr.2016.238 http://dx.doi.org/10.1109/cvpr.2016.238
LOU Y H , BAI Y , LIU J , et al . VERI-wild: a large dataset and a new method for vehicle re-identification in the wild [C]// Proceedings of the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition . Long Beach : IEEE , 2019 : 3230 - 3238 . doi: 10.1109/cvpr.2019.00335 http://dx.doi.org/10.1109/cvpr.2019.00335
ZHOUY Y , SHAO L . Viewpoint-aware attentive multi-view inference for vehicle re-identification [C]// Proceedings of the 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition . Salt Lake City : IEEE , 2018 : 6489 - 6498 . doi: 10.1109/cvpr.2018.00679 http://dx.doi.org/10.1109/cvpr.2018.00679
LIU X B , ZHANG S L , HUANG Q M , et al . RAM: a region-aware deep model for vehicle re-identification [C]// Proceedings of the 2018 IEEE International Conference on Multimedia and Expo (ICME) . San Diego : IEEE , 2018 : 1 - 6 . doi: 10.1109/icme.2018.8486589 http://dx.doi.org/10.1109/icme.2018.8486589
WANG Z D , TANG L M , LIU X H , et al . Orientation invariant feature embedding and spatial temporal regularization for vehicle re-identification [C]// Proceedings of the 2017 IEEE International Conference on Computer Vision . Venice : IEEE , 2017 : 379 - 387 . doi: 10.1109/iccv.2017.49 http://dx.doi.org/10.1109/iccv.2017.49
MENG D H , LI L , LIU X J , et al . Parsing-based view-aware embedding network for vehicle re-identification [C]// Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition . Seattle : IEEE , 2020 : 7101 - 7110 . doi: 10.1109/cvpr42600.2020.00713 http://dx.doi.org/10.1109/cvpr42600.2020.00713
QIAN J J , JIANG W , LUO H , et al . Stripe-based and attribute-aware network: a two-branch deep model for vehicle re-identification [J]. Measurement Science and Technology , 2020 , 31 ( 9 ): 095401 . doi: 10.1088/1361-6501/ab8b81 http://dx.doi.org/10.1088/1361-6501/ab8b81
涂明君 , 林国军 , 冯凯 , 等 . 基于多机制融合增强的跨模态行人重识别方法 [J]. 液晶与显示 , 2026 , 41 ( 5 ): 675 - 686 .
TU M J , LIN G J , FENG K , et al . Cross-modal pedestrian re-identification method enhanced by multi-mechanism fusion [J]. Chinese Journal of Liquid Crystals and Displays , 2026 , 41 ( 5 ): 675 - 686 . (in Chinese)
寇旗旗 , 乔鑫 , 牛凯凯 , 等 . 关系建模和频谱变换的跨模态行人重识别方法 [J]. 光学 精密工程 , 2026 , 34 ( 6 ): 973 - 989 .
KOU Q Q , QIAO X , NIU K K , et al . Cross-modal person re-identification method relation modeling and spectral transformation [J]. Optics and Precision Engineering , 2026 , 34 ( 6 ): 973 - 989 . (in Chinese)
LIU X C , LIU W , MEI T , et al . A deep learning-based approach to progressive vehicle re-identification for urban surveillance [C]// Proceedings of the 14th European Conference on Computer Vision—ECCV 2016 . Amsterdam : Springer , 2016 : 869 - 884 .
姬晓飞 , 赵帅 , 宋京浩 , 等 . 基于姿势估计和特征融合的行人重识别算法 [J]. 电子测量与仪器学报 , 2024 , 38 ( 4 ): 187 - 194 .
JI X F , ZHAO S , SONG J H , et al . Person re-identification algorithm based on pose estimation and feature fusion [J]. Journal of Electronic Measurement and Instrumentation , 2024 , 38 ( 4 ): 187 - 194 . (in Chinese)
WANG K Q , SONG L , WANG C T , et al . On the use of deep learning for phase recovery [J]. Light: Science & Applications , 2024 , 13 ( 1 ): 4 . doi: 10.1038/s41377-023-01340-x http://dx.doi.org/10.1038/s41377-023-01340-x
LIU H S , WANG F , JIN Y , et al . Learning-based real-time imaging through dynamic scattering media [J]. Light: Science & Applications , 2024 , 13 ( 10 ): 194 . doi: 10.1038/s41377-024-01569-0 http://dx.doi.org/10.1038/s41377-024-01569-0
DONG Y B , BAI Y C , ZHANG Q M , et al . High-throughput optical neuromorphic graphic processing at millions of images per second [J]. eLight , 2025 , 5 ( 1 ): 29 . doi: 10.1186/s43593-025-00106-9 http://dx.doi.org/10.1186/s43593-025-00106-9
DUN X , ZHANG J , CHEN F S , et al . Neural array meta-imaging [J]. eLight , 2025 , 5 ( 1 ): 33 . doi: 10.1186/s43593-025-00107-8 http://dx.doi.org/10.1186/s43593-025-00107-8
JEONG H , YOON C , LIM H , et al . MT-former: multi-task hybrid transformer and deep support vector data description to detect novel anomalies during semiconductor manufacturing [J]. Light: Advanced Manufacturing , 2025 , 6 ( 2 ): 32 . doi: 10.37188/lam.2025.032 http://dx.doi.org/10.37188/lam.2025.032
周颖 , 裴盛虎 , 陈海永 , 等 . 基于多尺度自适应注意力的图像超分辨率网络 [J]. 光学 精密工程 , 2024 , 32 ( 6 ): 843 - 856 . doi: 10.37188/ope.20243206.0843 http://dx.doi.org/10.37188/ope.20243206.0843
ZHOU Y , PEI S H , CHEN H Y , et al . Image super-resolution network based on multi-scale adaptive attention [J]. Optics and Precision Engineering , 2024 , 32 ( 6 ): 843 - 856 . (in Chinese) . doi: 10.37188/ope.20243206.0843 http://dx.doi.org/10.37188/ope.20243206.0843
滕嘉玮 , 田蕾 , 江山 , 等 . 卫星关键部件智能识别与三维重建方法 [J]. 中国光学(中英文) , 2025 , 18 ( 6 ): 1376 - 1387 . doi: 10.37188/CO.2025-0091 http://dx.doi.org/10.37188/CO.2025-0091
TENG J W , TIAN L , JIANG S , et al . Intelligent recognition and 3D reconstruction method for satellite key components [J]. Chinese Optics , 2025 , 18 ( 6 ): 1376 - 1387 . (in Chinese) . doi: 10.37188/CO.2025-0091 http://dx.doi.org/10.37188/CO.2025-0091
黄庆东 , 苏宇辉 , 刘依华 , 等 . 基于PATN和自注意力机制的多分辨率人体姿态迁移方法 [J]. 液晶与显示 , 2025 , 40 ( 9 ): 1381 - 1389 . doi: 10.37188/cjlcd.2025-0100 http://dx.doi.org/10.37188/cjlcd.2025-0100
HUANG Q D , SU Y H , LIU Y H , et al . Multi-resolution human posture transfer algorithm based on PATN and self-attention mechanism [J]. Chinese Journal of Liquid Crystals and Displays , 2025 , 40 ( 9 ): 1381 - 1389 . (in Chinese) . doi: 10.37188/cjlcd.2025-0100 http://dx.doi.org/10.37188/cjlcd.2025-0100
DOSOVITSKIY A , BEYER L , KOLESNIKOV A , et al . An image is worth 16×16 words: transformers for image recognition at scale [C]. International Conference on Learning Representations . Online : OpenReview.net , 2021 .
HE S T , LUO H , WANG P C , et al . TransReID: transformer-based object re-identification [C]// Proceedings of the 2021 IEEE/CVF International Conference on Computer Vision . Montreal : IEEE , 2021 : 14993 - 15002 . doi: 10.1109/iccv48922.2021.01474 http://dx.doi.org/10.1109/iccv48922.2021.01474
GONG R , ZHANG X , PAN J A , et al . Vehicle reidentification based on convolution and vision transformer feature fusion [J]. IEEE MultiMedia , 2024 , 31 ( 2 ): 61 - 68 . doi: 10.1109/mmul.2024.3398189 http://dx.doi.org/10.1109/mmul.2024.3398189
QIU M , CHRISTOPHER L A , CHIEN S , et al . Adaptive aspect ratios with patch-mixup-ViT-based vehicle ReID [C]// Proceedings of the 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) . Hyderabad : IEEE , 2025 : 1 - 5 . doi: 10.1109/icassp49660.2025.10890258 http://dx.doi.org/10.1109/icassp49660.2025.10890258
ZHAI Y , GUO X , LU Y , et al . In defense of the classification loss for person re-identification [C]// Proceedings of the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops . Long Beach : IEEE , 2019 : 1526 - 1535 . doi: 10.1109/cvprw.2019.00194 http://dx.doi.org/10.1109/cvprw.2019.00194
KHORRAMSHAHI P , KUMAR A , PERI N , et al . A dual-path model with adaptive attention for vehicle re-identification [C]// Proceedings of the 2019 IEEE/CVF International Conference on Computer Vision . Seoul : IEEE , 2019 : 6132 - 6140 . doi: 10.1109/iccv.2019.00623 http://dx.doi.org/10.1109/iccv.2019.00623
JIN X , LAN C L , ZENG W J , et al . Uncertainty-aware multi-shot knowledge distillation for image-based object re-identification [C]// Proceedings of the 34th AAAI Conference on Artificial Intelligence . New York : AAAI Press , 2020 : 11165 - 11172 . doi: 10.1609/aaai.v34i07.6774 http://dx.doi.org/10.1609/aaai.v34i07.6774
KHORRAMSHAHI P , PERI N , CHEN J C , et al . The devil is in the details: self-supervised attention for vehicle re-identification [C]// Proceedings of the 16th European Conference on Computer Vision—ECCV 2020 . Glasgow : Springer , 2020 : 369 - 386 . doi: 10.1007/978-3-030-58568-6_22 http://dx.doi.org/10.1007/978-3-030-58568-6_22
SHEN F , XIE Y , ZHU J Q , et al . GiT: graph interactive transformer for vehicle re-identification [J]. IEEE Transactions on Image Processing , 2023 , 32 : 1039 - 1051 . doi: 10.1109/tip.2023.3238642 http://dx.doi.org/10.1109/tip.2023.3238642
LI S Y , SUN L , LI Q L . CLIP-ReID: exploiting vision-language model for image re-identification without concrete text labels [C]// Proceedings of the 37th AAAI Conference on Artificial Intelligence . Washington : AAAI Press , 2023 : 1405 - 1413 . doi: 10.1609/aaai.v37i1.25225 http://dx.doi.org/10.1609/aaai.v37i1.25225
ZHANG X , LING Y N , LI K G , et al . Multimodality adaptive transformer and mutual learning for unsupervised domain adaptation vehicle re-identification [J]. IEEE Transactions on Intelligent Transportation Systems , 2024 , 25 ( 12 ): 20215 - 20226 .
HE Y L , WANG Z C , CHEN J , et al . EvoPrompt-ReID: a bilevel optimization framework for prompt-encoder co-evolution in image re-identification [C]// Proceedings of the 2026 IEEE/CVF Conference on Computer Vision and Pattern Recognition . Denver : IEEE , 2026 : 6758 - 6767 .
ZHU J Q , ZENG H Q , HUANG J C , et al . Vehicle re-identification using quadruple directional deep learning features [J]. IEEE Transactions on Intelligent Transportation Systems , 2020 , 21 ( 1 ): 410 - 420 . doi: 10.1109/tits.2019.2901312 http://dx.doi.org/10.1109/tits.2019.2901312
SUN K , PANG X Y , ZHENG M F , et al . Heterogeneous context interaction network for vehicle re-identification [J]. Neural Networks , 2024 , 169 : 293 - 306 . doi: 10.1016/j.neunet.2023.10.032 http://dx.doi.org/10.1016/j.neunet.2023.10.032
ZHU J Y , PARK T , ISOLA P , et al . Unpaired image-to-image translation using cycle-consistent adversarial networks [C]// Proceedings of the 2017 IEEE International Conference on Computer Vision . Venice : IEEE , 2017 : 2242 - 2251 . doi: 10.1109/iccv.2017.244 http://dx.doi.org/10.1109/iccv.2017.244
XU Y , RONG L L , ZHOU X L , et al . Cross-domain evaluation for vehicle re-identification [C]// Proceedings of the 2021 IEEE 6th International Conference on Cloud Computing and Big Data Analytics (ICCCBDA) . Chengdu : IEEE , 2021 : 474 - 477 . doi: 10.1109/icccbda51879.2021.9442574 http://dx.doi.org/10.1109/icccbda51879.2021.9442574
PENG J J , WANG H B , XU F Q , et al . Cross domain knowledge learning with dual-branch adversarial network for vehicle re-identification [J]. Neurocomputing , 2020 , 401 : 133 - 144 . doi: 10.1016/j.neucom.2020.02.112 http://dx.doi.org/10.1016/j.neucom.2020.02.112
0
Views
43
下载量
0
CSCD
Publicity Resources
Related Articles
Related Author
Related Institution
京公网安备11010802024621