Image editing method, network training method, related device and electronic equipment

An image editing and image technology, applied in image data processing, graphic image conversion, neural learning methods, etc., can solve the problem of image artifacts, distortion, etc., and achieve the effect of improving the output image quality

Active Publication Date: 2021-01-05
BEIJING BAIDU NETCOM SCI & TECH CO LTD
View PDF9 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] At present, the image editing method usually converts the image texture or appearance through the convolutional neural network. However, the output image after editing is easy to generate artifacts and distortion parts.

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
  • Image editing method, network training method, related device and electronic equipment
  • Image editing method, network training method, related device and electronic equipment
  • Image editing method, network training method, related device and electronic equipment

Examples

Experimental program
Comparison scheme
Effect test

no. 1 example

[0037] like figure 1 As shown, the present application provides an image editing method, comprising the following steps:

[0038] Step S101: Acquire a first image.

[0039] In this embodiment, the image editing method relates to the field of artificial intelligence, specifically to the field of computer vision technology and deep learning technology, which can be applied to electronic equipment, and the electronic equipment can be a server or a terminal, which is not specifically limited here.

[0040] The first image may be an image collected in real time, may also be a pre-stored image, may also be a picture sent by other devices, or may also be a picture obtained from a network.

[0041] For example, a device such as a mobile phone or a computer can be used to capture an image in real time and edit the image, or obtain an image that was previously captured and stored in the device, and edit the image, or receive an image from other devices Send an image and perform image ...

no. 2 example

[0076] like figure 2 As shown, the present application provides a network training method, comprising the following steps:

[0077] Step S201: Obtain a training sample image; wherein, the training sample image includes a training input image and a training output image, the training input image includes a first image content, and the training output image includes a second image content;

[0078] Step S202: Input the training sample image into a cycle-consistent GAN; wherein, the cycle-consistency GAN includes a first GAN and a second GAN, and the first GAN includes a second GAN A generator, the first generator includes a first spontaneous motion module, and the first spontaneous motion module is used to transform the first image content according to a first geometric transformation relationship to generate a first target image, so The first spontaneous motion module is further configured to transform the second image content edited based on the second generative confrontati...

no. 3 example

[0113] like Figure 4 As shown, the present application provides an image editing device 400, including:

[0114] A first acquiring module 401, configured to acquire a first image;

[0115] The first input module 402 is configured to input the first image to the trained cycle consistency generation confrontation network; wherein, the cycle consistency generation confrontation network includes a first generator, and the first generator includes a second generator A spontaneous motion module, the first spontaneous motion module is used to transform the image content to be edited in the first image according to the trained first geometric transformation relationship to generate a second image, and the second image includes the transformed image content;

[0116] An output module 403, configured to output a third image based on the second image generated by the first spontaneous movement module.

[0117] Optionally, the first generator further includes a first attention module ...

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 image editing method, a network training method, a related device and electronic equipment, which relate to the technical field of artificial intelligence such as computer vision and deep learning. According to the specific implementation scheme, a first image is acquired; the first image is inputted into a trained cyclic consistency generative adversarial network; the cyclic consistency generative adversarial network comprises a first generator, the first generator comprises a first spontaneous motion module, and the first spontaneous motion module is used for transforming to-be-edited image content in the first image according to a trained first geometric transformation relationship so as to generate a second image; the second image comprises the converted image content; and a third image is outputted based on the second image generated by the first spontaneous motion module. According to the technology of the invention, the problem that the output image quality is relatively low in the image editing technology is solved, and the output image quality of image editing is improved.

Description

technical field [0001] This application relates to the field of artificial intelligence, in particular to the field of computer vision and deep learning technology, and in particular to an image editing method, a network training method, related devices and electronic equipment. Background technique [0002] Image editing technology refers to a technology that edits a given image to generate an edited image in the target domain. This technology can be widely used in style transfer, sketch-to-photo conversion, and tag-based image synthesis. And face editing, etc. [0003] At present, the image editing method is usually to convert the image texture or appearance through the convolutional neural network. However, the output image after editing is easy to generate artifacts and distortion parts. Contents of the invention [0004] The disclosure provides an image editing method, a network training method, related devices and electronic equipment. [0005] According to a first...

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): G06T11/00G06T3/00G06K9/62G06N3/04G06N3/08G06N20/20
CPCG06T11/008G06T3/0012G06N3/08G06N20/20G06N3/045G06F18/214
Inventor 何声一洪智滨刘家铭胡天舒马明明郭汉奇
Owner BEIJING BAIDU NETCOM SCI & TECH CO LTD
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