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

A face attribute classification method based on multi-layer depth feature information

A technology of attribute classification and deep feature, applied in the field of computer vision, can solve the problem that the effect cannot be practically applied.

Active Publication Date: 2020-01-10
PCI TECH GRP CO LTD
View PDF2 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] Since the effects obtained by common methods cannot be applied in practice, the industry urgently needs a specific method that can accurately analyze the attributes of faces and apply them to actual scenes.

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 attribute classification method based on multi-layer depth feature information
  • A face attribute classification method based on multi-layer depth feature information
  • A face attribute classification method based on multi-layer depth feature information

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0016] A face attribute classification method based on multi-layer depth feature information, the specific steps are as follows:

[0017] S1: x is a face image from any angle, y is a front face image, find f so that f(x)=y, Construct multilayer f like this i = θ(w, x), so that f holds true. Here, the w parameter is learned through deep learning, so as to find the f function. First, by preprocessing the front face image, images with different angles are rotated as training images, and the corresponding positive face The face image is the desired result. In order to make the input and output of the network an image of the same size, the feature layer is followed by an upsampling layer. The loss function uses L2 norm to compare the last feature layer with the front face image. Through step-by-step iterative optimization, the The last feature layer is close to the front face image, and the finally trained network is the f to be found. Through this function, the input face image ...

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 attribute classification method based on multi-layer depth feature information, the specific steps are as follows: S1: Assume that x is a face image at any angle, S2: Use a local feature area extraction algorithm to transform the original image into local information , S3: Use a multi-layer feature extraction algorithm to extract multi-layer feature information, S4: Use a multi-feature fusion algorithm to adaptively fuse multi-dimensional information. The invention provides a face attribute classification technology based on multi-layer depth information. The method can effectively identify the attributes of people in face images, thereby realizing intelligent video monitoring and intelligent judgment. The method uses deep learning to analyze human faces The attribute samples are used for training, and the network model is divided into a public convolution layer, a feature extraction layer, and an attribute classification layer. In the feature extraction layer, connecting multi-layer features and fusing information of multiple scales helps to extract more distinguishable features.

Description

technical field [0001] The invention relates to the technical field of computer vision, in particular to a face attribute classification method based on multi-layer depth feature information. Background technique [0002] Intelligent video surveillance is based on digital and networked video surveillance, but it is different from general networked video surveillance. It is a higher-end video surveillance application. Smart video surveillance systems are able to identify different objects. Find abnormalities in the monitoring screen, and can issue alarms and provide useful information in the fastest and best way, so as to assist security personnel to deal with crises more effectively, and minimize false positives and false positives. Face attribute analysis in intelligent video surveillance is the key technology to realize this link. At present, the commonly used face attribute analysis methods are roughly divided into two categories. One is to use traditional feature extra...

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): G06K9/00G06N3/04
CPCG06V40/168G06V40/172G06N3/045
Inventor 丁保剑冯琰一王洋
Owner PCI TECH GRP 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