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

An image classification method based on octal convolution neural network

A convolutional neural network and classification method technology, applied in biological neural network models, neural architectures, instruments, etc., can solve problems such as affecting the accuracy of human behavior recognition, ignoring spatial dependencies, and destroying the color characteristics of real environments.

Active Publication Date: 2018-12-25
SOUTHEAST UNIV
View PDF2 Cites 13 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Traditional CNN is only suitable for the feature extraction of grayscale images or color image sub-channels, ignoring the spatial dependence between channels, destroying the color features of the real environment, thus affecting the accuracy of human behavior recognition

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 octal convolution neural network
  • An image classification method based on octal convolution neural network
  • An image classification method based on octal convolution neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0061] The present invention will be further described in detail below in conjunction with the drawings:

[0062] figure 1 It is a flowchart of the present invention, which mainly includes the following steps:

[0063] Step 1: Input N training images with a size of m×n pixels, preprocess the N training images, and learn the imaginary part of the octon. N, m, and n are positive integers, and N can be divisible by 8. Embed the octonion X into the real value and express the octonion with an eight-order real matrix, and apply the addition and multiplication of the matrix to the addition and multiplication of the octonion.

[0064] First, perform a real batch normalization operation on the input feature map R, and output the data of the i-1th layer of the network O i-1 Expressed as a four-dimensional matrix (m, f, p, q), where m is the size of a batch of data, f is the number of feature maps, p, q are the width and height of the feature maps, if you consider each feature map as It is a f...

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 an image classification method based on an octal convolution neural network. Firstly, a training image is input and is expressed as an octal matrix. Secondly, the octuple convolution neural network is established and trained to get the network parameters of each layer, that is, the training model. Then, the calibration image set is used to calibrate and adjust to the optimal network parameters. Finally, the test images are tested, the classification results are counted, and the recognition rate is calculated. The octal convolutional neural network constructed by the invention preserves the internal structure of the image by using the octal matrix expression mode, so that in various classification tasks, the constructed network can obtain higher classification accuracy rate of the image compared with the traditional method.

Description

Technical field [0001] The invention relates to an image classification method based on an oction convolutional neural network, which belongs to the technical field of deep learning. Background technique [0002] Deep Learning (Deep Learning: DL) is a new machine learning structure proposed by Professor Hinton of the University of Toronto in Canada in the international authoritative journal "Science" in 2006, which integrates an unsupervised layer-wise pretraining structure. It is effectively combined with the structure of Deep Neural Networks (DNN). Deep learning technology has attracted widespread attention from academia and industry, and has made breakthroughs in speech recognition, image recognition, and medical assisted diagnosis. The construction, promotion and reasonable explanation of deep learning network is one of the important research contents of the current basic theory of artificial intelligence application. In 1998, LeCun et al. proposed the classic LeNet-5 two-d...

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): G06K9/62G06N3/04
CPCG06N3/045G06F18/24G06F18/214
Inventor 伍家松徐玲孔佑勇杨冠羽章品正杨淳沨姜龙玉舒华忠
Owner SOUTHEAST 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