IEICE TRANSACTIONS on Fundamentals
Vehicle Re-Identification Based on Quadratic Split Architecture and Auxiliary Information Embedding
Summary :
Convolutional neural network (CNN) based vehicle re-identificatioin (ReID) inevitably has many disadvantages, such as information loss caused by downsampling operation. Therefore we propose a vision transformer (Vit) based vehicle ReID method to solve this problem. To improve the feature representation of vision transformer and make full use of additional vehicle information, the following methods are presented. (I) We propose a Quadratic Split Architecture (QSA) to learn both global and local features. More precisely, we split an image into many patches as “global part” and further split them into smaller sub-patches as “local part”. Features of both global and local part will be aggregated to enhance the representation ability. (II) The Auxiliary Information Embedding (AIE) is proposed to improve the robustness of the model by plugging a learnable camera/viewpoint embedding into Vit. Experimental results on several benchmarks indicate that our method is superior to many advanced vehicle ReID methods.
- Publication
- IEICE TRANSACTIONS on Fundamentals Vol.E105-A No.12 pp.1621-1625
- Publication Date
- 2022/12/01
- Publicized
- 2022/05/24
- Online ISSN
- 1745-1337
- DOI
- 10.1587/transfun.2022EAL2008
- Type of Manuscript
- LETTER
- Category
- Image
Authors
Tongwei LU
Hubei Key Laboratory of Intelligent Robot (Wuhan Institute of Technology),Wuhan Institute of Technology
Hao ZHANG
Hubei Key Laboratory of Intelligent Robot (Wuhan Institute of Technology),Wuhan Institute of Technology
Feng MIN
Hubei Key Laboratory of Intelligent Robot (Wuhan Institute of Technology),Wuhan Institute of Technology
Shihai JIA
Hubei Key Laboratory of Intelligent Robot (Wuhan Institute of Technology),Wuhan Institute of Technology
Keyword
Latest Issue
Copyrights notice of machine-translated contents
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Cite this
Copy
Tongwei LU, Hao ZHANG, Feng MIN, Shihai JIA, "Vehicle Re-Identification Based on Quadratic Split Architecture and Auxiliary Information Embedding" in IEICE TRANSACTIONS on Fundamentals,
vol. E105-A, no. 12, pp. 1621-1625, December 2022, doi: 10.1587/transfun.2022EAL2008.
Abstract: Convolutional neural network (CNN) based vehicle re-identificatioin (ReID) inevitably has many disadvantages, such as information loss caused by downsampling operation. Therefore we propose a vision transformer (Vit) based vehicle ReID method to solve this problem. To improve the feature representation of vision transformer and make full use of additional vehicle information, the following methods are presented. (I) We propose a Quadratic Split Architecture (QSA) to learn both global and local features. More precisely, we split an image into many patches as “global part” and further split them into smaller sub-patches as “local part”. Features of both global and local part will be aggregated to enhance the representation ability. (II) The Auxiliary Information Embedding (AIE) is proposed to improve the robustness of the model by plugging a learnable camera/viewpoint embedding into Vit. Experimental results on several benchmarks indicate that our method is superior to many advanced vehicle ReID methods.
URL: https://global.ieice.org/en_transactions/fundamentals/10.1587/transfun.2022EAL2008/_p
Copy
@ARTICLE{e105-a_12_1621,
author={Tongwei LU, Hao ZHANG, Feng MIN, Shihai JIA, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={Vehicle Re-Identification Based on Quadratic Split Architecture and Auxiliary Information Embedding},
year={2022},
volume={E105-A},
number={12},
pages={1621-1625},
abstract={Convolutional neural network (CNN) based vehicle re-identificatioin (ReID) inevitably has many disadvantages, such as information loss caused by downsampling operation. Therefore we propose a vision transformer (Vit) based vehicle ReID method to solve this problem. To improve the feature representation of vision transformer and make full use of additional vehicle information, the following methods are presented. (I) We propose a Quadratic Split Architecture (QSA) to learn both global and local features. More precisely, we split an image into many patches as “global part” and further split them into smaller sub-patches as “local part”. Features of both global and local part will be aggregated to enhance the representation ability. (II) The Auxiliary Information Embedding (AIE) is proposed to improve the robustness of the model by plugging a learnable camera/viewpoint embedding into Vit. Experimental results on several benchmarks indicate that our method is superior to many advanced vehicle ReID methods.},
keywords={},
doi={10.1587/transfun.2022EAL2008},
ISSN={1745-1337},
month={December},}
Copy
TY - JOUR
TI - Vehicle Re-Identification Based on Quadratic Split Architecture and Auxiliary Information Embedding
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 1621
EP - 1625
AU - Tongwei LU
AU - Hao ZHANG
AU - Feng MIN
AU - Shihai JIA
PY - 2022
DO - 10.1587/transfun.2022EAL2008
JO - IEICE TRANSACTIONS on Fundamentals
SN - 1745-1337
VL - E105-A
IS - 12
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - December 2022
AB - Convolutional neural network (CNN) based vehicle re-identificatioin (ReID) inevitably has many disadvantages, such as information loss caused by downsampling operation. Therefore we propose a vision transformer (Vit) based vehicle ReID method to solve this problem. To improve the feature representation of vision transformer and make full use of additional vehicle information, the following methods are presented. (I) We propose a Quadratic Split Architecture (QSA) to learn both global and local features. More precisely, we split an image into many patches as “global part” and further split them into smaller sub-patches as “local part”. Features of both global and local part will be aggregated to enhance the representation ability. (II) The Auxiliary Information Embedding (AIE) is proposed to improve the robustness of the model by plugging a learnable camera/viewpoint embedding into Vit. Experimental results on several benchmarks indicate that our method is superior to many advanced vehicle ReID methods.
ER -
English
