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

An image classification method based on concise unsupervised convolutional network

An unsupervised, convolutional network technology, applied in the field of image processing and deep learning, can solve the problems of high complexity of deep convolutional neural network models, strict requirements for labeled image data, and difficulty in training network models. The effect of geometric invariance, improved robustness, and improved accuracy

Active Publication Date: 2018-06-19
UNIV OF ELECTRONICS SCI & TECH OF CHINA
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] In order to overcome the problems of high complexity of traditional deep convolutional neural network models, large number of parameters, difficulty in training network models, and strict requirements for labeled image data, this invention studies how to use a simple unsupervised algorithm to reduce The complexity of the network model, at the same time, a large number of unlabeled images can be used to train the network model

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
  • An image classification method based on concise unsupervised convolutional network
  • An image classification method based on concise unsupervised convolutional network
  • An image classification method based on concise unsupervised convolutional network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0029] The specific implementation steps adopted by the present invention to solve its technical problems are as follows:

[0030] Step 1: The training image set Each training picture in is divided into multiple image blocks of size w×h, and the pixel composition dimension of each image block is R M The vector, where M=w×h×d, d represents the channel value of the image, for RGB pictures, d=3, for grayscale pictures, d=1; the entire training image set contains T image blocks in total, all These T image block vectors form a matrix P={p 1 ,...,p t ,...,p T}, where, t=1,...,T,p t ∈ R M ;

[0031] Step 2: Preprocessing the T image blocks;

[0032] Normalize according to formula (1), and whiten according to formula (2)(3)(4):

[0033]

[0034]

[0035]

[0036]

[0037] Among them, mean( ) is to find the mean value of the vector, var( ) is to find the variance of the vector, cov( ) is to find the covariance matrix of the vector, Eig( ) is to find the eigenvalue ...

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 an image classification method based on a simple and unsupervised convolutional network, which belongs to the technical field of image processing and deep learning. The present invention utilizes the classic unsupervised clustering algorithm K-means to cluster the image blocks of the training image set, and each cluster center obtained is the convolution kernel in the network model, and abandons the traditional convolution network repeatedly passing through The stochastic gradient descent algorithm is used to obtain the time-consuming process of the convolution kernel; in addition, the present invention enhances the robustness of the network to image deformation by proposing a probability pooling method. Through the simple unsupervised deep convolutional network classification model proposed by the present invention, the model training time can be effectively reduced, and the recognition ability of the model for changing scene pictures can be improved at the same time.

Description

technical field [0001] The invention belongs to the technical field of image processing and deep learning, and relates to efficient image classification processing, in particular to an implementation scheme of image classification based on a concise unsupervised convolutional network. Background technique [0002] In recent years, image classification has received extensive attention and applications in the fields of industry, manufacturing, military, and medical treatment. Although its development situation is very good, as the coverage of practical applications gradually expands, massive image data will follow. Both the scale of the image database and the diversity of image content have reached an unprecedented peak. This makes the traditional Image processing methods are overwhelmed. Faced with such a large amount of image information, how to accurately classify images has become a research hotspot in related fields today. [0003] In the field of pattern recognition, t...

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
CPCG06V30/194G06F18/2411
Inventor 董乐张宁
Owner UNIV OF ELECTRONICS SCI & TECH OF CHINA
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