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

Face identification method based on cosine similarity measure learning

A technology of cosine similarity and metric learning, applied in the field of face recognition, can solve problems such as inappropriate matching of face features, achieve the effect of compact space and improve recognition accuracy

Inactive Publication Date: 2017-01-11
江苏华通晟云科技有限公司
View PDF3 Cites 22 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

These measurement methods are designed by humans and do not take advantage of the information between face samples, such as the connection and difference between samples from the same person and samples from different people, so it is not suitable for face feature matching directly.

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 identification method based on cosine similarity measure learning
  • Face identification method based on cosine similarity measure learning
  • Face identification method based on cosine similarity measure learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0029] Such as figure 1 Shown, a kind of face recognition method based on cosine similarity measure learning of the present invention comprises the following steps:

[0030] (1) For any input image, first detect faces.

[0031] (2) If a face is detected, continue to locate the feature points of the face, such as figure 2 shown. Correct the human face to the in-plane level according to the human eye coordinates, and crop the image proportionally.

[0032] (3) Extract facial features. In this example, high-dimensional and multi-scale LBP features are used. The specific steps are as follows:

[0033] (3-1) Normalize face images to 5 scales, namely 300, 200, 150, 100 and 75.

[0034] (3-2) For each feature point, take the surrounding 40×40 area and divide it into non-overlapping 4×4 blocks, and calculate the LBP of the uniform mode in each block (the feature dimension is 59). In this example, 16 feature points are taken for each scale, so the face feature dimension is 75520=...

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 identification method based on cosine similarity measure learning. The method is characterized in that the method comprises the following steps: carrying out face detection, feature point positioning and normalization and feature extraction on a face image; calculating all training samples and verification samples to obtain feature vectors; estimating an optimum transfer matrix A through a cosine similarity measure learning method; calculating similarity between the samples, and comparing the similarity with a preset threshold value; and if the similarity is larger than the preset threshold value, judging the face images are from the same person, or otherwise, judging the face images are not from the same person. The method is simple to calculate, can improve generalization capability of the existing measure, and enhances face identification precision.

Description

technical field [0001] The invention relates to a face recognition method, in particular to a face recognition method based on cosine similarity measure learning. Background technique [0002] Face recognition is a human biometric technology, which has practical and potential applications in many fields, such as criminal investigation, document verification, security monitoring, etc. It is precisely because of the broad application prospects of face recognition that it has increasingly become a hot spot in the field of pattern recognition and artificial intelligence. [0003] Face recognition is easily affected by various factors, such as: lighting, expression, angle, sample size, etc. At present, how to improve the accuracy of face recognition is mainly reflected in how to obtain effective identification features. In recent years, with the continuous research of deep learning, people can learn good features from a large number of samples. In terms of matching, especially...

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
CPCG06V40/161G06V40/168G06F18/217
Inventor 武克杰鲁星星吴建伟
Owner 江苏华通晟云科技有限公司
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