Unlock instant, AI-driven research and patent intelligence for your innovation.

A face detection method based on deep learning

A technology of face detection and deep learning, applied in the field of face recognition, which can solve problems such as high time complexity, slow recognition speed, and complex implementation, and achieve the effect of simple network structure, fast extraction speed, and good robustness

Active Publication Date: 2020-05-22
青岛格莱瑞智能控制技术有限公司
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] Among them, the eigenface method has the advantages of simplicity, speed, and practicability, but because it relies too much on the image gray-level correlation between the training sample set and the test sample set in actual application, and requires the test sample to be relatively close to the training sample, Therefore, there are great limitations and cannot be widely used in practice.
The elastic graph matching method has good recognition accuracy and good applicability, but its disadvantages are also obvious: high time complexity, slow recognition speed, and complicated implementation

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
  • A face detection method based on deep learning
  • A face detection method based on deep learning
  • A face detection method based on deep learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0021] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0022] The face detection method based on deep learning in this embodiment includes the following three steps:

[0023] Step 1. Perform face pair alignment on the collected face images; face alignment is the front-end processing of face recognition. By scaling, rotating, cutting and other operations on the face images, all face images can be standardized according to a certain standard . Face alignment corrects factors that are not conducive to recognition such as large differences in postures and exaggerated expressions during the acquisition process. In this specific embodiment, the existing CFAN face alignment algorithm is used for face alignment, which is a coarse-to-fine autoencoder network to solve complex nonlinear mapping processes. Such as figure 1 , CFAN consists of 4 stacked autoencoder networks (SAN), each autoencoder network has a four-l...

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 detection method based on deep learning, comprising steps: step 1, face alignment of collected face images; step 2, performing deep convolution on the face images processed by step 1 Neural network extracts face features; step 3, normalize the feature vector output in step 2, and then perform cosine similarity calculation with the feature vector originally stored in the system, and finally compare the threshold value to obtain the Face detection results. The face detection method based on deep learning of the present invention has the advantages of fast extraction of face image features, high accuracy, good robustness, simple network structure and the like.

Description

technical field [0001] The invention relates to the technical field of face recognition, in particular to a face feature recognition method suitable for multiple tasks. Background technique [0002] The general flow of the face recognition method is as follows: the system input is generally one or a series of face images containing undetermined identities, and several face images with known identities in the face database, and its output is a series of A similarity score indicating the identity of the face to be recognized. The face recognition method mainly includes three parts: feature extraction, feature matching, and similarity calculation. The similarity obtained by different calculations needs to be compared with the expected threshold threshold. When the similarity is equal to or higher than the threshold, the identity of the face is determined. Therefore, the quality of the feature extraction algorithm in the face recognition algorithm directly affects the accuracy...

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
Patent Type & Authority Patents(China)
IPC IPC(8): G06K9/00G06K9/62G06N3/08
CPCG06N3/08G06V40/168G06F18/22
Inventor 宋永端刘秀兰刘永杨琳赖俊峰李攀飞张子涛张云福
Owner 青岛格莱瑞智能控制技术有限公司