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

Image style converting party and device, apparatus, and storage medium

A style conversion and image technology, applied in the image field, can solve problems such as face edge deformation, no face photo optimization, and face skin color inconsistency, etc.

Active Publication Date: 2019-02-05
SHENZHEN SENSETIME TECH CO LTD
View PDF5 Cites 12 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The current technology does not optimize the style conversion of face photos. For example, when the existing methods are applied to selfie images, the common disadvantages are: the deformation of the edges of the face and the inconsistent skin color of the face caused by the image style conversion

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 style converting party and device, apparatus, and storage medium
  • Image style converting party and device, apparatus, and storage medium
  • Image style converting party and device, apparatus, and storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0129] The process of generating a style map using a neural network method is generally as follows: use a neural network model such as the VGG16 model or VGG19 to perform image feature extraction on an original image (Content Image) and a style image (Style Image), namely The content features are extracted from the original image, and the style features are extracted from the style image. By using the content features and style features to construct a loss function, calculate the loss value of a randomly initialized image and feed back the redrawn image to obtain a generated image (Generated Image), this generated image will be similar to the original image in content, but in style above will be similar to the style image. However, every time this algorithm generates an image, it needs to be trained once, which takes a long time.

[0130] Based on the fast style transfer algorithm, training a network can convert any image into the corresponding style of the network, so each 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 embodiment of the invention discloses an image style converting party, a device, a storage medium, wherein, the method comprises the following steps: obtaining an initial image to be converted into a style; Inputting a gradient of the initial image to an image style conversion model, and obtaining a feature map of the initial image in a gradient domain from the image style conversion model; The image style conversion model is obtained based on pixel-level loss and perceptual loss training in the gradient domain. According to the characteristic image of the initial image in the gradient domain, the image is reconstructed to obtain a style image.

Description

technical field [0001] The present invention relates to image technology, in particular to an image style conversion method, device, equipment, and storage medium. Background technique [0002] Image style transfer based on deep learning is a new research problem in recent years. Although the problem of image style conversion has always existed, it was only in 2015 that the German researcher Gatys used the method of neural network for the first time to open the door to create image art style with deep learning. The current technology does not optimize the style conversion of face photos. For example, when the existing methods are applied to selfie images, the common disadvantages are: the deformation of the edges of the face and the inconsistent skin color of the face caused by the image style conversion . Contents of the invention [0003] In view of this, the embodiments of the present invention provide an image style conversion method, device, device, and storage medi...

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): G06T3/00G06T7/90
CPCG06T7/90G06T2207/10024G06T3/04G06T11/001G06V40/171G06V10/454G06V10/82
Inventor 贺高远柳一村陈晓濠任思捷
Owner SHENZHEN SENSETIME TECH 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