.png){kind=icon}
Return
Pedestrian attribute recognition based on attribute correlation
.png){kind=icon}
PRE
AI
DOI：10.1007/s00530-022-00893-y.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Pedestrian attribute recognition is widely used in pedestrian tracking and pedestrian re-identification. This task confronts two fundamental challenges. One comes from its multi-label nature; the other one comes from the characteristics of data samples, such as the class imbalance and the partial occlusion. In this work, we propose a Cross Attribute and Feature Network (CAFN) that fully exploits the correlations between any pair of attributes for the pedestrian attribute recognition to tackle these challenges. Concretely, CAFN contains two modules: Cross-attribute Attention Module (C2AM) and Cross-feature Attention Module (CFAM). C2AM enables the network to automatically learn the relation matrix during the training process which can quantify the correlations between any pair of attributes in the attribute set, and CFAM is introduced to fuse different attribute features to generate more accurate and robust attribute features. Extensive experiments demonstrate that the proposed CAFN performs favorably compared with state-of-the-art approaches.
Keywords:
Pedestrian attribute recognition
Attention mechanism
Multi-label classification
Attribute correlation
Visual feature correlation
Journal
IF:
3.1
Papers:
2.8K
Citations:
2.7K
Organization
Cited Papers
ShareSaveShareSaveProperties of the double well potential and relaxation processes in a model glass.png)
PhysChemComm
PhysChemCommIF0
PREAI
ShareSaveShareSaveThermal comfort, perceived air quality, and cognitive performance when personally controlled air movement is used by tropically acclimatized personsIndoor Air
OAAI
Indoor AirIF0
OAAIShareSaveSignificance of nailfold videocapillaroscopy in patients with idiopathic inflammatory myopathiesRheumatologyOAAI
RheumatologyIF0
OAAIShareSave
View more