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

A zero-shot image classification method based on relative attribute random forest

A random forest and sample image technology, applied in the field of pattern recognition, can solve problems such as unreasonable, maximum likelihood estimation method error, image classification accuracy, etc.

Active Publication Date: 2019-03-01
中国科学院电子学研究所苏州研究院
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, this method has certain disadvantages: (1) It is unreasonable to assume that all known and unknown class images obey the Gaussian distribution; , so it will be affected by human subjective factors and the accuracy of the model is not high; (3) There is a large error in the maximum likelihood estimation method, which will also affect the accuracy of image classification

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 zero-shot image classification method based on relative attribute random forest
  • A zero-shot image classification method based on relative attribute random forest
  • A zero-shot image classification method based on relative attribute random forest

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0061] Embodiments of the present invention are described in detail below, examples of which are shown in the drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the figures are exemplary only for explaining the present invention and should not be construed as limiting the present invention.

[0062] Zero-shot image classification methods based on relative attribute random forests, such as image 3 shown, including the following steps:

[0063] Step 1: If Figure 4 In (1), given the underlying features of the known class image and the class label set {x 1 ,x 2 ,...,x S ;y 1 ,y 2 ,...,y S}, the underlying feature set of unknown class images {z 1 ,z 2 ,...,z U}, the ordered attribute pair set {O 1 ,...,O M}, the set of similar attribute pairs of known class images {S 1 ,...,S M}, the number T of random trees and the s...

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 present invention proposes a zero-sample image classification method based on a relative attribute random forest, which establishes an attribute ranking score model for images of unknown categories according to the relative relationship between image categories and image attributes, and uses the attribute ranking score models of all images as training samples To train a random forest classifier, and finally predict the label of the test image according to the attribute ranking score of the test image and the trained random forest classifier. The method of the invention can realize zero-sample image classification, and has the advantages of high classification recognition rate, strong model stability and the like.

Description

technical field [0001] The invention belongs to the field of pattern recognition, in particular to a zero-sample image classification method based on relative attribute random forest. Background technique [0002] 1.1 Zero-shot image classification [0003] Zero-shot image classification is one of the research hotspots in the field of pattern recognition at present. Different from traditional image classification problems, the images classified and recognized by zero-shot image classification in the test phase are not involved in the training of the classifier model. Such as figure 1 As shown, the marked images in the training phase cover the three categories of "Lion", "Athletic shoes" and "Polar bear" (that is, known categories), while the images in the testing phase appear in the category of "Stiletto" (that is, unknown categories ), since the "Stiletto" category did not participate in the training of the classifier, the classifier will not be able to predict its label...

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): G06K9/62G06K9/66
CPCG06V30/194G06F18/2415
Inventor 乔雪彭晨段贺刘久云胡岩峰刘振
Owner 中国科学院电子学研究所苏州研究院
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