Unlock instant, AI-driven research and patent intelligence for your innovation.

Image style migration method based on self-attention mechanism mapping

An attention and image technology, applied in the field of image processing, can solve the problem of not being able to effectively capture the structure of the graph

Pending Publication Date: 2022-04-12
CHONGQING UNIV OF POSTS & TELECOMM
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, most methods only apply low-level image features, and cannot effectively capture the graph structure, so they cannot effectively perform style transfer.

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 migration method based on self-attention mechanism mapping
  • Image style migration method based on self-attention mechanism mapping
  • Image style migration method based on self-attention mechanism mapping

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0043] Embodiments of the present invention are described below through specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific implementation modes, and various modifications or changes can be made to the details in this specification based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the diagrams provided in the following embodiments are only schematically illustrating the basic concept of the present invention, and the following embodiments and the features in the embodiments can be combined with each other in the case of no conflict.

[0044] Wherein, the accompanying drawings are for illustrative purposes only, and represent only schematic diagrams, rather than physical drawings, and should...

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 image style migration method based on self-attention mechanism mapping, and belongs to the field of image processing. Firstly, feature extraction is performed on a content image and a style image, respective content features and style features are extracted respectively, and then a self-attention mechanism is utilized to perform mapping processing on the content features of the content image and the content of the style image. And finally, combining the mapped content features with the style features of the style image to synthesize a target image. According to the method, the mapping between the content features of the style image and the content image is completed by utilizing self-attention, the problems of local style and overall migration effect in the existing image style migration process are solved, and the method has reliability and robustness.

Description

technical field [0001] The invention belongs to the field of image processing, and relates to a method for image style transfer based on self-attention mechanism mapping. Background technique [0002] Image style transfer refers to given a content map and a style map, using a machine learning algorithm to render the content map into a picture with the artistic style of the style map, while ensuring that the structure of the content map does not change. [0003] Gatys et al. proposed a new method that can flexibly extract the style of any style image and render its style into any content graph in an iterative manner. However, the method of Gatys et al. is through image iteration. , a white noise image can be rendered into a target image through repeated iterations. This process is very slow and time-consuming. Justin et al. proposed a fast migration algorithm. By iteratively training the network model, after the network training is completed, it only needs to be performed on...

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): G06T3/00G06K9/62G06N3/04G06N3/08G06V10/40G06V10/774G06V10/82
Inventor 刘伯红张龙
Owner CHONGQING UNIV OF POSTS & TELECOMM