Pedestrian multi-attribute identification method combining local region detection and multi-level feature capture

Abstract

The invention provides a pedestrian multi-attribute identification method combining local region detection and multi-level feature capture, and relates to the technical field of pedestrian attribute identification. According to the pedestrian multi-attribute identification method combining local region detection and multi-level feature capture, a pedestrian segmentation module, a feature fusion module and a multi-task learning module are included, and the pedestrian segmentation module, the feature fusion module and the multi-task learning module are fused into an end-to-end framework. The invention provides the pedestrian multi-attribute identification method combining local region detection and multi-level feature capture, which is based on deep learning of an identity identification technology, reduces irrelevant information, makes full use of attribute correlation information, solves the problem of unbalanced sample distribution, enhances the identification capability of pedestrian attributes, establishes a complete pedestrian multi-attribute identification framework, solves the problem of environmental information interference, and can enhance the identification capability of each local attribute by fully utilizing multi-attribute correlation and additional auxiliary information.

Description

technical field [0001] The invention relates to the technical field of pedestrian attribute recognition, in particular to a pedestrian multi-attribute recognition method combining local area detection and multi-level feature capture. Background technique [0002] Pedestrian attributes refer to the attributes of people, such as face, clothes, accessories, age, etc. Accurate identification of these attributes can not only improve the ability of intelligent machines to understand humans, but also play a key role in many practical application technologies, such as: video-based Intelligent commercial recommendation, pedestrian re-identification in video surveillance, and attribute-based pedestrian retrieval. [0003] The existing pedestrian attribute recognition methods mainly include methods based on manual feature extraction and methods based on deep learning. The methods based on manual feature extraction mainly use low-level features such as color and texture for identificati...

AI Technical Summary

Problems solved by technology

[0007] Aiming at the deficiencies of the prior art, the present invention provides a pedestrian multi-attribute recognition method combining local area detection and multi-level feature capture, which solves the problem that the method based on manual feature extraction does not have strong generalization, and the feature extraction It takes a long time and is not suitable for feature extraction of large data sets. The mainstream pedestrian attribute recognition method based on deep learning still cannot effectively use the correlation between attributes, and the attribute recognition effect for unbalanced sample distribution is poor, and the environment The impact will also interfere with the effect of the attention mechanism on the enhancement of local attribute area features

Images