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

Multi-label Image Classification Method Based on Manifold Learning and Gradient Boosting Model

A technology of manifold learning and classification method, applied in the field of two-stage partial multi-label learning, can solve problems such as limited prediction performance, and achieve the effect of improving prediction performance, excellent performance, and realizing disambiguation

Active Publication Date: 2022-04-12
ZHEJIANG UNIV
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Existing PML methods all focus on one aspect of candidate label set disambiguation or label correlation extraction, so these methods have limited predictive performance

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
  • Multi-label Image Classification Method Based on Manifold Learning and Gradient Boosting Model
  • Multi-label Image Classification Method Based on Manifold Learning and Gradient Boosting Model
  • Multi-label Image Classification Method Based on Manifold Learning and Gradient Boosting Model

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0064] The technical solution of the present invention will be further described in conjunction with specific implementation and examples.

[0065] Such as figure 1 , specific embodiments of the present invention and its implementation process are as follows:

[0066] In the first stage, label disambiguation is performed first, including step 1 and step 2:

[0067] step 1:

[0068] First, from the pre-specified training data set Build a weighted graph in Among them, V represents the collection of picture feature vectors, V={x i |1≤i≤n},x i Represents the feature vector of the i-th picture, i represents the ordinal number of the picture, and n represents the training data set The total number of pictures in; E represents the set of connections between every two pictures, E={(x i ,x j )|i≠j,x j ∈kNN(x i )}, kNN(x i ) represents the feature vector x to the i-th picture i The set of feature vectors of the first k pictures with the closest Euclidean distance, (x i ...

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 a method for classifying pictures with multiple labels based on manifold learning and gradient lifting model. Construct a weighted graph from the training data set, obtain the non-negative weight matrix by solving the above first minimization model, establish the second minimization model according to the weighted graph and solve it to obtain the reconstructed label matrix, and transform the training data set according to the reconstructed label matrix Construct and train a binary correlation model to predict the label matrix; establish a regressor to minimize the eigenvector matrix of the image, use the iterative prediction result matrix to enhance the eigenvector matrix, combine the negative gradient matrix to construct a data set and train and learn to obtain a weak regressor , and sum all weak regressors to obtain the final regressor, which is used to process and judge the image to be tested. The present invention can make full use of the correlation between the multi-label data of the picture to improve the multi-label classification prediction performance of the picture, realize the disambiguation of the partial label data, improve the accuracy and robustness, and its performance is better than the existing The image is too much label method.

Description

technical field [0001] The invention relates to the problem of over-many labels in label classification, in particular to a two-stage over-many label learning method based on manifold learning and gradient promotion. Background technique [0002] In the image multi-label classification problem, a picture can be associated with multiple labels at the same time. A common assumption in traditional image multi-label classification problems is that each image in the training dataset is accurately labeled, that is, supervised. Unfortunately, in many real image multi-label classification problems, noise-free labels are difficult to obtain. In contrast, obtaining a set of candidate labels is very easy. This type of problem is defined as PartialMulti-Label Learning (PML) for pictures. [0003] The basic assumption of PML is that the correct label of the image is hidden in the set of candidate labels, and it is invisible to the learner. The most intuitive approach to PML is to con...

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): G06V10/764G06K9/62
CPCG06F18/2411
Inventor 陈刚强宇周王皓波谌晨陈珂胡天磊寿黎但伍赛
Owner ZHEJIANG UNIV
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