Intelligent face editing method, device, storage medium and equipment

An intelligent human and editing technology, applied in the field of intelligent face editing, can solve the problems of unable to generate face, unable to edit the overall appearance of the face, etc.

Active Publication Date: 2021-09-03
中科计算技术创新研究院
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, existing techniques cannot edit the overall appearance of a face, nor can they generate realistic and realistic faces when pure sketches are used as input.

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
  • Intelligent face editing method, device, storage medium and equipment
  • Intelligent face editing method, device, storage medium and equipment
  • Intelligent face editing method, device, storage medium and equipment

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0056] like figure 1 As shown, the present embodiment uses image decoupling techniques, using local to global methods, a smart face editing method is proposed.

[0057] In this example, the face is divided into 5 parts: left eye, right eye, nose, mouth, background, after completing the image block, design local decoupling module separately decouple features and generation of images of each part, respectively. After obtaining the result of each portion, the overall fusion module will be fused to give the result of a global consistency.

[0058] The network structure of the present embodiment includes five local decoupling modules for decoupling geometry and appearance information, and one global fusion module achieves a combination of local features and generates high quality, with global consistency. During the process of network training, use the exchanged strategy, and design the loop consistent feature constraint to ensure robustness and generalization of the network frame.

[...

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 an intelligent face editing method, device, storage medium and equipment. Applicable to the fields of computer vision and computer graphics. Provided are an intelligent face editing method, device, storage medium and equipment. The technical solution adopted by the present invention is: an intelligent face editing method, which is characterized in that: the geometric feature image of the face and the feature image of the appearance of the face are input into the corresponding and trained local decoupling module according to the face position, Extract the geometric features and appearance features corresponding to each part of the face; the local decoupling module generates local images and local intermediate feature maps corresponding to each part of the face based on the geometric features and appearance features; through the trained global fusion module, each part of the face is The corresponding local intermediate feature maps are fused to generate the final face image with the geometric and appearance features.

Description

Technical field [0001] The present invention relates to a smart face editing method, a device, a storage medium, and a device. Suitable for computer vision, computer graphics field. Background technique [0002] Face synthesis technology is one of the important topics in the field of digital image processing, and has many related technologies around high-quality face synthesis. Based on deep learning face synthesis techniques mainly include the following two categories: through Gaussian sampling, use the generated counterfeit network (GaN), synthesizing new faces; use conditional generated anti-network, input semantic tag, sketch, attribute tag, etc. Information, synthesizing the corresponding face. Sketch synthesis of real people's faces although there are many technologies, most of them cannot control the appearance of the generated face, or the face synthesis is poor. [0003] For face editing, some existing technologies use label face attribute marker data to decouple the GaN...

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): G06T11/00G06T5/50
CPCG06T5/50G06T11/00G06T2207/20081G06T2207/20221G06T2207/30201
Inventor 高林陈姝宇刘锋林
Owner 中科计算技术创新研究院
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