Natural image target material visual feature mapping method based on generative adversarial network

A technology of visual features and natural images, applied in biological neural network models, image enhancement, image analysis, etc., can solve the problems of material visual feature learning and its mapping methods, and achieve the effect of wide application prospects

Inactive Publication Date: 2018-11-23
LANZHOU UNIVERSITY OF TECHNOLOGY
View PDF4 Cites 14 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0002] With the continuous development of computer technology, feature-based mapping change methods have become a popular research direction. This type of method tra

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
  • Natural image target material visual feature mapping method based on generative adversarial network
  • Natural image target material visual feature mapping method based on generative adversarial network
  • Natural image target material visual feature mapping method based on generative adversarial network

Examples

Experimental program
Comparison scheme
Effect test

Example Embodiment

[0057] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments, but this embodiment described by the accompanying drawings is exemplary, only used to explain the present invention, and cannot limit the scope of the present invention.

[0058] figure 1 The overall framework of the natural image target material visual feature mapping method based on generative adversarial network is given, in which the solid line represents the forward cycle generation process: X→Y, the dotted line represents the backward cycle generation process: Y→X, the dotted line represents Parameter update.

[0059] In this paper, a method for visual feature mapping of natural image target material based on generative adversarial network is invented. The main steps are as follows:

[0060] 1. The establishment of the data sample set

[0061] Images with different visual feature information of materials in natural scenes are collected fr...

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 provides a natural image target material visual feature mapping method based on a generative adversarial network. A deeply unsupervised learning way is used for learning the unlabeled natural image target material visual feature to obtain the high-order expression of image target material visual feature space, a mapping network about the material visual feature space between a sourcedomain image and a target domain image is learnt and established, the material visual feature of the source domain image is mapped to the material visual feature of the target domain, so that the target domain image has the material visual feature information of the source domain image, and finally, an image of which the material visual feature is mapped is obtained. The method learns from the unlabeled natural image to obtain the material visual feature information, the task target of which the material visual feature is mapped among different images is carried out, a corresponding solutionis put forward by aiming at the task target, a good result is obtained, and the method has important theoretical significance and practical values.

Description

technical field [0001] The invention relates to the fields of image processing and deep learning technology, computer vision and artificial intelligence, in particular to a method for visual feature mapping of natural image target materials based on a generative confrontation network. Background technique [0002] With the continuous development of computer technology, the feature-based mapping change method has become a popular research direction. This kind of method transforms some features in one space into other space features through mapping, but for material visual features Learning and its mapping methods are less studied. The material visual feature is a high-level comprehensive feature based on a visual mechanism, which describes the material feature information of the target corresponding to the image or image area. properties and their properties. Based on the above analysis, the present invention proposes a natural image target material visual feature mapping m...

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
IPC IPC(8): G06N3/08G06N3/04G06T7/194
CPCG06N3/088G06T7/194G06T2207/10004G06T2207/20084G06T2207/20081G06N3/045
Inventor 李策贾盛泽万玉奇张栋刘昊张亚超蓝天
Owner LANZHOU UNIVERSITY OF TECHNOLOGY
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