Face recognition method based on migration layered network

A face recognition and network technology, applied in the field of computer vision, can solve problems such as excellent models, inability to obtain, and inability to obtain enough labeled training samples

Inactive Publication Date: 2018-09-28
NANJING UNIV
View PDF4 Cites 28 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, in some scenarios, due to cost constraints such as resources and manpower, it is impossible ...

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 migration layered network
  • Face recognition method based on migration layered network
  • Face recognition method based on migration layered network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0089] The present invention adopts the above-mentioned scheme and achieves satisfactory effects on the standard face data set Yale, AR. The Yale data set was collected and created by the Computational Vision and Control Center of Yale University. There are 165 pictures of 15 people from various angles in the data set, including the influence of various visual factors, such as changes in lighting, expression and posture. AR: Including 120 people, more than 4,000 images, corresponding to human faces under different expressions and lighting conditions. It is a highly recognized database.

[0090] The specific implementation is as follows:

[0091] Step 1: The initial models with general significance that can be selected are mainly the ImageNet classification model and the VGG model. The present invention is to perform face recognition, so the VGG-Face model is selected.

[0092] Step 2: Build a hierarchical network for face classification, divide the convolutional layer into 5 groups...

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 face recognition method based on a migration layered network. The method comprises the steps of S1, selecting a pre-training model; S2, establishing the layered network for face classification; S3, determining fine-adjustment network layers according to the selected pre-training model and the similar degree of existing face data; S4, for a face training data set, carryingout preprocessing operation, comprising color enhancement, rotation, translation, and addition of random noises; S5, training the existing data through utilization of a caffe depth learning library;and S6, carrying out face recognition through utilization of the trained model.

Description

Technical field [0001] The invention belongs to the field of computer vision, and particularly relates to a face recognition method based on a migration hierarchical network. Background technique [0002] In recent years, deep learning has been widely used in the field of computer vision. Deep learning methods have used more training sets than traditional methods and adopted a deeper network structure, and have also made significant progress in face recognition. However, in some scenarios, due to cost constraints such as resources and manpower, it is impossible to obtain enough labeled training samples for training, resulting in a failure to obtain a better model. Summary of the invention [0003] Purpose of the invention: The present invention proposes a face recognition scheme based on a migration hierarchical network that can be used when there is less face data. Use a trained model with universal significance, and use the weight of this model as the pre-value of the new mode...

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): G06K9/00G06N3/04G06N3/08
CPCG06N3/08G06V40/172G06N3/045
Inventor 杨育彬甘元柱李瑮朱瑞
Owner NANJING UNIV
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