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

Face age synthesis method based on style fusion and domain selection structure

A style, face image technology, applied in the field of face generation, can solve the problems of face identity information loss, training instability, etc.

Pending Publication Date: 2022-01-04
HEBEI UNIV OF TECH
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The present invention can effectively solve the problems of face identity information loss and training instability

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 age synthesis method based on style fusion and domain selection structure
  • Face age synthesis method based on style fusion and domain selection structure
  • Face age synthesis method based on style fusion and domain selection structure

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0146] The face age synthesis method based on style fusion and domain selection structure of the present embodiment, the specific steps are as follows:

[0147] In the first step, the multi-scale features of the input face image are extracted through the encoder network:

[0148] In step 1.1, the face image is normalized, quantized, and the image resolution is adjusted to 256×256 to obtain a preprocessed face image.

[0149] In step 1.2, input the preprocessed face image into the first layer of convolution module, and obtain the output feature of the first layer encoder as

[0150] In step 1.3, the output features of the first layer encoder are extracted through the encoder network to extract depth features of different scales, and the specific operation is shown in formula (1);

[0151]

[0152] in, Represents the output feature of the i-th layer encoder, Conv represents the convolution module, N enc Indicates the number of encoder network layers;

[0153] This comp...

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 relates to a face age synthesis method based on style fusion and a domain selection structure, and the method comprises the following steps: carrying out preprocessing of a face public data set, enabling each face sample to correspond to an age label, and setting the number of age domains; constructing an adversarial network of style fusion and multi-domain discrimination, wherein the adversarial network comprises a generator network based on style fusion and a domain selection discriminator network; the domain selection discriminator network comprises a plurality of domain selection structures and a full connection layer, and each domain selection structure is composed of two types of functions: a basic function and a plurality of domain functions; for each batch of input images, only a basic function and a specific domain function are used to extract features, wherein the number of domain functions is consistent with the number of age domains. According to the method, the problems of human face identity information loss and unstable training can be effectively solved.

Description

technical field [0001] The technical solution of the present invention relates to the field of human face generation, specifically a method for synthesizing human face age based on style fusion and domain selection structure. Background technique [0002] Face age synthesis, that is, face age aging / rejuvenation, aims to generate face appearances of different ages for a given face image, while generating images that also preserve the identity of the source image. In recent years, with the rapid development of deep learning, a major breakthrough has been made in face age synthesis, which can not only provide technical support for data augmentation for face recognition systems, but also assist criminal cases by predicting the future appearance of lost children and suspects. It can also be applied to the production of special effects for film and television entertainment. [0003] The existing face age synthesis methods can be divided into two categories, which are the synthesi...

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 Applications(China)
IPC IPC(8): G06T3/00G06N3/08G06N3/04
CPCG06N3/084G06N3/045G06T3/02G06T3/04
Inventor 郭迎春夏伟毅于洋朱叶阎刚郝小可师硕刘依吕华于明
Owner HEBEI UNIV OF TECH
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