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

A Correlation Feedback Method Based on Extreme Learning Machine

An extreme learning machine and related feedback technology, which is applied to computer components, instruments, calculations, etc., can solve the problem of reduced processing time and achieve the effects of improved accuracy, high processing speed, and good feedback effects

Active Publication Date: 2019-02-01
BEIJING UNIV OF TECH
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

It can be seen from the experiment that the method using the Gaussian kernel has a higher classification accuracy, and at the same time, the processing time has also decreased to a large extent.

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
  • A Correlation Feedback Method Based on Extreme Learning Machine
  • A Correlation Feedback Method Based on Extreme Learning Machine
  • A Correlation Feedback Method Based on Extreme Learning Machine

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0020] The present invention will be further described below in conjunction with specific embodiments.

[0021] The process of the relevant feedback method based on the ELM of the present invention is as follows: figure 1 shown, including the following steps:

[0022] Step 1, input a query image;

[0023] Step 2: Retrieve the image to obtain the retrieval result, and let the user mark the result;

[0024] Step 2.1, in the retrieval system, what the user needs to mark is a retrieval result or a feedback result. The images marked by the user are used as the dataset L, and the images that have not been marked by the user are used as the dataset U. At this time, the datasets that have not been marked by the user are the entire gallery datasets D-L. Positive examples P and negative examples N are included in the labeled dataset, so L=P∪N.

[0025] Step 3, extract SIFT features, Color features, and LBP features from the marked images respectively;

[0026] Step 4, using three k...

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

A related feedback method based on an extreme learning machine. The invention inputs a query image; retrieves the image to obtain a retrieval result, and allows the user to mark the result; respectively extracts SIFT features, Color features, and LBP features; use three types of features to train three basic classifiers; put the images in the search gallery into the three basic classifiers, vote according to the prediction results, and automatically mark each unlabeled picture; retrain and update the classification device; classify gallery images; return results. The present invention is based on the extreme learning machine, by introducing human query intentions, performing human-computer interaction, and effectively using unmarked gallery images to enrich learning data, which can greatly improve the accuracy of image feedback and greatly improve the processing speed. Good control makes the expression of the image in the computer more in line with the understanding of the semantics of the image by human beings, so that the present invention has a good feedback effect.

Description

technical field [0001] The invention belongs to the field of image processing, and relates to a related feedback method in image retrieval, in particular to a related feedback method based on an ultra-limited learning machine. Background technique [0002] The current society has entered the era of huge data mainly based on multimedia information data, among which digital image information data is the most prominent. Compared with other multimedia data, image data is richer in content and more intuitive in expression, which has inevitably and has become a very important form of information sharing in people's daily life. In the face of increasing image information data, the huge amount of information hidden in the image data can be effectively mined, so that the image information actually needed by the user can be quickly and accurately found in the large-scale image database. This trend has gradually become a major research topic in related fields such as computer vision a...

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): G06F16/58G06K9/62
CPCG06F16/5838G06F18/241
Inventor 段立娟董帅马伟杨震赵则明
Owner BEIJING UNIV OF TECH