Contour symmetry constraint generative adversarial network-based multi-pose face recognition method

A face recognition and multi-pose technology, applied in character and pattern recognition, instruments, computer components, etc., can solve problems such as fitting process and calculation process difficulties, and achieve the effect of meeting high-precision requirements

Active Publication Date: 2018-07-27
GUILIN UNIV OF ELECTRONIC TECH
View PDF6 Cites 25 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

This type of method requires a lot of depth information, and

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
  • Contour symmetry constraint generative adversarial network-based multi-pose face recognition method
  • Contour symmetry constraint generative adversarial network-based multi-pose face recognition method
  • Contour symmetry constraint generative adversarial network-based multi-pose face recognition method

Examples

Experimental program
Comparison scheme
Effect test

Example Embodiment

[0029] Examples:

[0030] Reference figure 1 , figure 2 , A multi-pose face recognition method based on contour symmetry constraint generative confrontation network, including the following steps:

[0031] 1) Data preprocessing: In order to establish better face template features, the multi-pose face database is divided into training images and test images, and the training images and test images are normalized. This example is in Multi- The PIE face image database verifies the effectiveness of this technical solution. The database contains 754,204 face images of 337 individuals with different poses, lighting, and expressions. In this example, a sub-image set containing -75° -11 poses in the +75° angle range, with 15° intervals between poses. The selected face data are all normal expressions under normal lighting. Set the image size alignment and cropping in the database to 64x64, and the first 210 people As the training image, the remaining 127 people are used as the test image...

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 contour symmetry constraint generative adversarial network-based multi-pose face recognition method. The method is characterized by comprising the following steps of 1) performing data preprocessing; 2) creating a contour constraint generative network; 3) creating a symmetry constraint adversarial network; 4) training and balancing the networks; and 5) performing reconstruction and recognition. According to the method, the pose angle deflection influence of a face image can be effectively eliminated; a characteristic that a face is robuster under multiple poses is extracted; especially global quality and local details are mutually constrained under large-angle pose reconstruction; contour characteristic information of the front face is kept; and the high-precisiondemand on the multi-pose face recognition in actual application can be met.

Description

Technical field [0001] The invention relates to the field of intelligent image processing and pattern recognition, and in particular to a multi-posture face recognition method based on Contour Symmetry Constraint Generative Adversarial Network (SC-GAN). Background technique [0002] Multi-pose face recognition (Multi-pose Face Recognition) is a hot spot of machine vision research in recent years, especially the proposal and rise of deep learning, which makes face recognition technology make significant progress and rapid development in many fields . However, in reality, face images are susceptible to various factors such as different environments, lighting, expressions, and postures, which affect the accuracy of face recognition. Among them, face posture is a very challenging problem. [0003] In order to solve the intra-class changes brought about by posture changes in face recognition, researchers have achieved certain results. At present, the main technologies are divided into ...

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): G06K9/00G06K9/62
CPCG06V40/172G06F18/21322G06F18/214
Inventor 欧阳宁刘力元林乐平莫建文袁华首照宇张彤陈利霞
Owner GUILIN UNIV OF ELECTRONIC TECH
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