Cross-domain face recognition algorithm, storage medium and processor

A face recognition, cross-domain technology, applied in the field of face recognition algorithms, can solve the problems of low recognition accuracy and low recognition efficiency

Pending Publication Date: 2020-07-07
SHENZHEN KUANG CHI SPACE TECH CO LTD
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

For cross-domain recognition, traditional recognition algorithms such as Facenet cann

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
  • Cross-domain face recognition algorithm, storage medium and processor
  • Cross-domain face recognition algorithm, storage medium and processor
  • Cross-domain face recognition algorithm, storage medium and processor

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0028] figure 1 It is a schematic diagram of the Facenet neural network structure in the prior art. Facenet neural network structure, the input module passes through module A, module B-1, module B-2, module C-1, module C-2, module D, and then through the Siftmax layer, and then the Triplet-Loss calculation is performed.

[0029] figure 2 It is a schematic diagram of the structure of the cross-domain face recognition algorithm module of the present invention. Such as figure 2 shown in figure 1 On the basis of the existing Facenet neural network, this embodiment adds an MMD module and a synthetic loss function module. S10 is a neural network module, S20 is a maximum difference mean module, and S30 is a loss function module. Module A to Module E are feature extraction network modules. Among them, module A is a conventional feature extraction network; modules B-1, C-1 and D draw on the idea of ​​Inception-Resnet to reduce the amount of parameter calculation while ensuring...

Embodiment 2

[0039] An embodiment of the present invention also provides a storage medium, which includes a stored program, wherein, when the above program is running, the above-mentioned flow of the face attribute recognition method is executed.

[0040] Optionally, in this embodiment, the above-mentioned storage medium may be configured to store program codes for executing the following flow of the face attribute recognition method:

[0041] S11, building a Facenet neural network;

[0042] S12, adding the average pooling layer and the Faltten layer at the highest dimension layer of the Facenet neural network feature vector, and changing the feature vector into a single-channel one-dimensional vector;

[0043] S13. Calculate the maximum mean difference loss with a single-channel one-dimensional vector;

[0044] S14. Add the maximum mean difference loss to the loss function of the Facenet neural network, and jointly participate in backpropagation and gradient derivation.

[0045] Optiona...

Embodiment 3

[0048] An embodiment of the present invention also provides a processor, the processor is used to run a program, wherein, when the program is running, the steps in the above-mentioned face attribute recognition method are executed.

[0049] Optionally, in this embodiment, the above program is used to perform the following steps:

[0050] S11, building a Facenet neural network;

[0051] S12, adding the average pooling layer and the Faltten layer at the highest dimension layer of the Facenet neural network feature vector, and changing the feature vector into a single-channel one-dimensional vector;

[0052] S13. Calculate the maximum mean difference loss with a single-channel one-dimensional vector;

[0053] S14. Add the maximum mean difference loss to the loss function of the Facenet neural network, and jointly participate in backpropagation and gradient derivation.

[0054] Optionally, for specific examples in this embodiment, reference may be made to the above-mentioned emb...

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 provides a cross-domain face recognition algorithm, a storage medium and a processor. The cross-domain face recognition algorithm method comprises the following steps of: establishing aFacenet neural network; adding an average pooling layer and a Faltten layer at the highest dimension layer of the feature vector of the Facenet neural network, and converting the feature vector into asingle-channel one-dimensional vector; calculating the maximum mean value difference loss by using a single-channel one-dimensional vector; and adding the maximum mean value difference loss into a loss function of the Facenet neural network, and making the maximum mean value difference loss and the loss function jointly participate in back propagation and gradient derivation. A network structureof an original Facenet algorithm is improved, MMD values of different domains are calculated at the highest layer of the feature dimension, and the MMD values are added into the synthetic loss function; due to the fact that the inter-domain statistical distribution difference is eliminated through the improved algorithm, a cross-domain face recognition effect is achieved.

Description

【Technical field】 [0001] The present invention relates to the technical field of face recognition algorithms, in particular to a cross-domain face recognition algorithm, a storage medium and a processor. 【Background technique】 [0002] Face recognition is a biometric technology for identification based on human facial feature information. A series of related technologies that use a video camera or camera to collect images or video streams containing human faces, automatically detect and track human faces in the images, and then perform facial recognition on the detected faces, usually also called portrait recognition and facial recognition. . [0003] Most of the existing face recognition algorithms can solve the problem of single-domain face recognition (that is, the image to be recognized and the training sample image have the same statistical distribution characteristics). The Facenet algorithm proposed by Florian Schroff et al. is currently a good single-domain face rec...

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): G06K9/00G06N3/04
CPCG06V40/16G06N3/045G06N3/04G06F18/00
Inventor 刘若鹏栾琳赵盟盟
Owner SHENZHEN KUANG CHI SPACE TECH CO LTD
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