System and method for synthesizing front face from side face in end-to-end manner based on conditional generative adversarial network

A conditional generation, end-to-end technology, applied in the field of image processing, can solve the problems of face recognition interference, poor quality, unsatisfactory face recognition effect, etc., and achieve the effect of simple and fast training

Pending Publication Date: 2020-03-17
SHANGHAI UNIV
View PDF3 Cites 8 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

In the actual shooting, it will be affected by external factors (such as noise, light, angle, etc.), resulting in a large number of faces with poor quality and non-frontal postures in the captured face images, which have brought difficulties to face recognition. A lot of interference, resulting in unsatisfactory face recognition, for example, misidentifying A in the image as B

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
  • System and method for synthesizing front face from side face in end-to-end manner based on conditional generative adversarial network
  • System and method for synthesizing front face from side face in end-to-end manner based on conditional generative adversarial network
  • System and method for synthesizing front face from side face in end-to-end manner based on conditional generative adversarial network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0056] Embodiment one: see figure 1 , figure 2 , image 3 and Figure 4 , a system implementation process of end-to-end face image generation based on conditional generative adversarial network of the present invention.

Embodiment 2

[0057] Embodiment 2: A method of end-to-end face image generation based on a conditional generative confrontation network of the present invention, such as Figure 1-Figure 4 shown, including the following steps:

[0058] (1) Input multiple preprocessed real face images to the generator, and the generator encodes, transcodes and decodes the input real face images to generate a composite image that fits the real image distribution, and converts multiple After preprocessing, the real face image and the synthetic image are input into the discriminator to obtain the real probability of the real face image and the real probability of the synthetic image, and iteratively update the parameters of the generator and the discriminator until they converge to determine the parameters of the generator and discriminator. the built model;

[0059] (2) Input the profile pose face image to be synthesized into the determined model, and obtain the generated front pose face image through a forwa...

Embodiment 3

[0082] Embodiment three: in Figure 5Among them, the preprocessing module (1), the generator module (2), and the discriminator module (3) are connected in series, and the data flows through the preprocessing module (1) to the generator module (2), and then to the discriminator module (3). Among them, the preprocessing module (1) is composed of two parts: the tailoring module (1-1) and the scaling module (1-2), and the serial data of the two flows from the tailoring module to the scaling module; the generator module (2) is composed of three parts: A decoding module (2-1), an encoding module (2-2), and a generating module (2-3), wherein the encoding module (2-2) and the generating module (2-3) are connected in parallel, and receive signals from the decoding module (2-3) simultaneously. 1) The given data; the discrimination module (3) consists of two parts: the convolution module (3-1) and the fully connected module (3-2), and the data flows from the convolution module (3-1) to t...

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 an end-to-end face image generation method based on a conditional generative adversarial network. The method comprises the following specific steps: inputting a plurality of preprocessed real face images to a generator, and carrying out encoding, carrying out transcoding and decoding processing on the input real face images by the generator to generate a composite image fitting real image distribution; inputting the synthesized image of the obtained real image distribution and a plurality of originally input preprocessed real face images into a discriminator to obtain the real probability of the real face images and the real probability of the synthesized image, and iteratively updating the parameters of the generator and the discriminator until the parameters converge so as to determine a model constructed by the generator and the discriminator; and inputting a to-be-synthesized side posture human face image into the model, and obtaining a generated front posture human face image through one-time forward transmission. The method is high in efficiency, can process extreme illumination conditions and multi-angle side posture transformation, and does not needto estimate a three-dimensional model or two-dimensional feature points of a human face.

Description

technical field [0001] The invention belongs to the technical field of image processing, and mainly relates to an end-to-end system and method for synthesizing a frontal face from a side face based on a conditional generative confrontation network. Background technique [0002] In recent years, face recognition technology has developed rapidly, and its application fields are becoming wider and wider. In the actual shooting, it will be affected by external factors (such as noise, light, angle, etc.), resulting in a large number of faces with poor quality and non-frontal postures in the captured face images, which have brought difficulties to face recognition. A large amount of interference causes unsatisfactory face recognition results, for example, A in the image is mistakenly identified as B. These problems will bring great inconvenience to the follow-up work. Therefore, how to generate a clear and recognizable high-quality face image of the frontal pose from the face imag...

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): G06K9/00G06K9/62G06N3/04
CPCG06V40/168G06N3/045G06F18/214
Inventor 方昱春李一帆袁秋龙涂小康
Owner SHANGHAI 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