Pedestrian re-identification data set enhancement method and system, terminal, camera and medium

A pedestrian re-identification and data set technology, applied in the field of image processing, can solve the problems that the generator and the discriminator are prone to fall into abnormal confrontation, the training of the generative confrontation network is unstable, and the training degree of freedom is large, so as to reduce the collection and labeling Cost, rich variety, and the effect of meeting training requirements

Pending Publication Date: 2021-05-14
深圳市网联安瑞网络科技有限公司
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0013] The training of the generated confrontation network is unstable, the training degree of freed

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 data set enhancement method and system, terminal, camera and medium
  • Pedestrian re-identification data set enhancement method and system, terminal, camera and medium
  • Pedestrian re-identification data set enhancement method and system, terminal, camera and medium

Examples

Experimental program
Comparison scheme
Effect test

Example Embodiment

[0070] Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, the same numerals in different drawings refer to the same or similar elements unless otherwise indicated. The implementations described in the following exemplary examples do not represent all implementations consistent with the present disclosure. Rather, they are merely examples of apparatuses and methods consistent with aspects of the present disclosure as recited in the appended claims.

[0071] like figure 1 As shown, the present invention is based on the pedestrian re-identification data set enhancement method of generating confrontation network, comprising the following steps:

[0072] S101: The network inputs three pictures of pedestrians, each of which is x i 、x j 、x k , where x j with x k It is the picture collected by the same pedestrian in different scenes...

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 data set enhancement method and system, a terminal, a camera and a medium, and relates to the technical field of image processing. The method comprises steps of inputting three pedestrian images through a network, and encoding by using an appearance encoder and a structure encoder respectively to obtain an appearance code and a structure code; using an Es encoder to encode xi, and obtaining a structure code Csi; using an Ea encoder to encode the xk to obtain an appearance code Cak of the xk, and judging by a discriminator D1 and a discriminator D2; iterating repeatedly to obtain an optimal generator; and generating a pedestrian data set by using the trained optimal generator, inputting three pedestrian pictures, generating pedestrian pictures with different appearances in a crossing manner, and generating pedestrian pictures of the same pedestrian ID with different postures, backgrounds, illumination and angles in a crossing manner. According to the invention, the pedestrian re-identification data set acquisition and labeling cost is reduced, the diversity of the pedestrian re-identification data set is enriched, and the development of the pedestrian re-identification technology is assisted.

Description

technical field [0001] The disclosure of the present invention relates to the technical field of image processing, and in particular to a pedestrian re-identification data set enhancement method, system, terminal, camera and medium. Background technique [0002] At present, geometric transformation includes enhancement methods such as flip, rotation, cropping, deformation, and scaling; common color transformation includes enhancement methods such as noise, blur, color transformation, erasure, and filling; synthesis and splicing enhancement methods; generation classes based on motion prediction Enhancement method. [0003] The defective of prior art is: [0004] Common methods of geometric transformation, color transformation, and synthesis are not suitable for pedestrian re-identification data enhancement; the diversity of samples generated by production methods based on action prediction is insufficient, such as pedestrian appearance, clothing, decoration, posture, backgro...

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
IPC IPC(8): G06K9/00G06K9/62G06N3/04G06N3/08
CPCG06N3/04G06N3/08G06V40/10G06F18/24G06F18/214
Inventor 廖伟石珺李志鹏廖勇杨阳朝易勇
Owner 深圳市网联安瑞网络科技有限公司
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