Geodesic
ALFormer: Attribute Localization Transformer in Pedestrian Attribute Recognition
Computer Science and Information Systems, Tome 21 (2024) no. 4.

Voir la notice de l'article provenant de la source Computer Science and Information Systems website

Pedestrian attribute recognition is an important task for intelligent video surveillance. However, existing methods struggle to accurately localize discriminative regions for each attribute. We propose Attribute Localization Transformer (ALFormer), a novel framework to improve spatial localization through two key components. First, we introduce Mask Contrast Learning (MCL) to suppress regional feature relevance, forcing the model to focus on intrinsic spatial areas for each attribute. Second, we design an Attribute Spatial Memory (ASM) module to generate reliable attention maps that capture inherent locations for each attribute. Extensive experiments on two benchmark datasets demonstrate state-of-the-art performance of ALFormer. Ablation studies and visualizations verify the effectiveness of the proposed modules in improving attribute localization. Our work provides a simple yet effective approach to exploit spatial consistency for enhanced pedestrian attribute recognition.
Keywords: spatial attention, attribute localization, contrast loss, random mask 
@article{CSIS_2024_21_4_a19,
     author = {Yuxin Liu and Mingzhe Wang and Chao Li and Shuoyan Liu},
     title = {ALFormer: {Attribute} {Localization} {Transformer} in {Pedestrian} {Attribute} {Recognition}},
     journal = {Computer Science and Information Systems},
     publisher = {mathdoc},
     volume = {21},
     number = {4},
     year = {2024},
     url = {http://geodesic.mathdoc.fr/item/CSIS_2024_21_4_a19/}
}
TY  - JOUR
AU  - Yuxin Liu
AU  - Mingzhe Wang
AU  - Chao Li
AU  - Shuoyan Liu
TI  - ALFormer: Attribute Localization Transformer in Pedestrian Attribute Recognition
JO  - Computer Science and Information Systems
PY  - 2024
VL  - 21
IS  - 4
PB  - mathdoc
UR  - http://geodesic.mathdoc.fr/item/CSIS_2024_21_4_a19/
ID  - CSIS_2024_21_4_a19
ER  - 
%0 Journal Article
%A Yuxin Liu
%A Mingzhe Wang
%A Chao Li
%A Shuoyan Liu
%T ALFormer: Attribute Localization Transformer in Pedestrian Attribute Recognition
%J Computer Science and Information Systems
%D 2024
%V 21
%N 4
%I mathdoc
%U http://geodesic.mathdoc.fr/item/CSIS_2024_21_4_a19/
%F CSIS_2024_21_4_a19
Yuxin Liu; Mingzhe Wang; Chao Li; Shuoyan Liu. ALFormer: Attribute Localization Transformer in Pedestrian Attribute Recognition. Computer Science and Information Systems, Tome 21 (2024) no. 4. http://geodesic.mathdoc.fr/item/CSIS_2024_21_4_a19/