Pedestrian Re-Identification Method Based on Sparse Attention Network

A recognition method and attention technology, applied in the field of computer vision, can solve the problem of loss of effective features and achieve the effect of avoiding loss

Active Publication Date: 2022-03-15
深圳万知达科技有限公司
View PDF9 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] What the present invention aims to solve is the problem that the existing pedestrian re-identification method loses a large number of effective features when performing deep learning, and provides a pedestrian re-identification method based on sparse attention network, which, under the condition of constant model complexity, It can significantly improve the extraction performance of detailed features of pedestrian images, alleviate the loss of effective features, and improve the accuracy of pedestrian classification

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Pedestrian Re-Identification Method Based on Sparse Attention Network
  • Pedestrian Re-Identification Method Based on Sparse Attention Network
  • Pedestrian Re-Identification Method Based on Sparse Attention Network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0051] In order to make the object, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with specific examples.

[0052] The pedestrian re-identification model constructed by the present invention is a sparse normalized compression-excitation network, such as figure 1 As shown, it is mainly composed of a backbone layer in the middle, 4 short connections on one side of the backbone layer, and 4 normalized compression-excitation modules on the other side of the backbone layer.

[0053] (1) Backbone layer:

[0054] The first convolutional layer, the convolutional layer is composed of filters with a kernel size of 7×7, which is used for dimensionality reduction. After dimensionality reduction, the image becomes 1 / 4 of the original image size, so this layer is mainly to reduce Calculations.

[0055] The second layer is the maximum pooling layer, that is, the maximum value is taken in th...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses a pedestrian re-identification method based on a sparse attention network. Firstly, the shallow features are transmitted to the deep features without loss through short connections; and features; then through the normalized compression-excitation module embedded in the backbone residual network to extract the detailed features of the image that are easy to be lost; finally, the above-mentioned features are multiplied, and finally the features obtained in the first part are added and sent to the Fully connected layer and classification regression layer to obtain classification and regression results. The sparse attention network of the present invention can effectively extract the detailed features of pedestrian photos of several pedestrian re-identification data sets.

Description

technical field [0001] The invention relates to the technical field of computer vision, in particular to a pedestrian re-identification method based on a sparse attention network. Background technique [0002] Pedestrian re-identification refers to re-confirming the identity of the same pedestrian in different monitoring scenes to make up for the visual limitations of a single camera. Pedestrian re-identification can be widely used in intelligent image understanding, intelligent video analysis, intelligent video detection and other fields. At present, the methods applied to pedestrian re-identification are mainly divided into: pedestrian re-identification based on artificial design features and pedestrian re-identification based on deep convolutional neural network. Pedestrian re-identification based on artificial design features mainly includes two parts: artificial design feature extraction and feature similarity measurement; pedestrian re-identification model based on de...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Patents(China)
IPC IPC(8): G06V40/10G06V10/77G06V10/774G06V10/764G06K9/62G06V10/82G06N3/04G06N3/08
CPCG06V40/10G06F18/2136G06F18/214G06F18/241
Inventor 张灿龙解盛李志欣
Owner 深圳万知达科技有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap