Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Face recognition method based on Transform and convolutional neural network

A convolutional neural network and face recognition technology, applied in biological neural network models, neural architecture, character and pattern recognition, etc., can solve the problems of limited receptive field and insufficient global information perception ability, to improve accuracy, increase The effect of model complexity

Pending Publication Date: 2022-08-02
SHANDONG ARTAPLAY INTELLIGENT TECH CO LTD
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The current face recognition method is to design a network model through a convolutional neural network. Although the CNN-based model has strong generalization ability and can effectively extract local information, this convolutional structure has limited perception of global information due to its limited receptive field. lack of ability

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
  • Face recognition method based on Transform and convolutional neural network
  • Face recognition method based on Transform and convolutional neural network
  • Face recognition method based on Transform and convolutional neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0035]In step a), the IResNet network has four Block modules, and the first 3×3 convolution operation in each Block module is retained, and the self-atteention module in the CoTNet network is replaced with the second 3×3 in each Block module. 3 convolution operations, the last of each Block module is combined with a SE module.

Embodiment 2

[0037] Preferably, the value of N in step b) is 112.

Embodiment 3

[0039] In step b), the Stem module is sequentially composed of a convolutional layer with a convolution kernel size of 3, a BN layer and a PReLu layer.

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

A human face recognition method based on Transform and a convolutional neural network comprises the steps of inputting a cut human face image into a human face recognition model so as to perform feature extraction on the human face image, performing feature matching after obtaining human face feature vectors of a fixed channel number, and taking human face features with high similarity as human face information of the same identity. The face recognition model is constructed on the basis of the most core self-attention mechanism in Transform in combination with a convolutional neural network architecture, and an SE (channel attention module) is introduced, so that the accuracy of face recognition is improved under the condition that the complexity of the model is not remarkably increased, and the application of a face recognition technology in an actual scene is facilitated.

Description

technical field [0001] The invention relates to the technical field of face recognition, in particular to a face recognition method based on Transformer and convolutional neural network Background technique [0002] With the rapid development of science and technology, more and more intelligent technologies have entered people's daily life. As a popular research topic with high social popularity and mature technology, face recognition is widely used in various fields. The current face recognition method uses the convolutional neural network to design the network model. Although the CNN-based model has strong generalization ability and can effectively extract local information, this convolutional structure has limited receptive field and cannot perceive global information. lack of ability. In natural scenes, the interference of posture changes, side faces and other factors will affect the accuracy of face recognition. SUMMARY OF THE INVENTION [0003] In order to overcom...

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): G06V40/16G06V10/40G06V10/74G06V10/764G06V10/82G06K9/62G06N3/04
CPCG06N3/045G06F18/22Y04S10/50
Inventor 刘永辉韩春港韩继泽杜浩谢恩鹏王志亮
Owner SHANDONG ARTAPLAY INTELLIGENT TECH CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com