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

Method and apparatus for increasing generalization capability of convolutional neural network

A convolutional neural network and generalization technology, which is applied in the field of improving the generalization ability of convolutional neural network, can solve the problems of reduced generalization ability of neural network, loss of data set information, and improvement of generalization ability of convolutional neural network model. To achieve the effect of improving the generalization ability

Inactive Publication Date: 2017-05-17
SHENZHEN INST OF ADVANCED TECH CHINESE ACAD OF SCI
View PDF0 Cites 42 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

When the training data presents an exponent distribution with a long tail, as the number of tail images increases, the generalization ability of the neural network decreases
[0005] Existing solutions usually use the method of discarding tail data, which causes the loss of information in the data set, and fundamentally does not make use of more sufficient inter-class information to help the convolutional neural network model improve the generalization ability

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
  • Method and apparatus for increasing generalization capability of convolutional neural network
  • Method and apparatus for increasing generalization capability of convolutional neural network
  • Method and apparatus for increasing generalization capability of convolutional neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0036] see figure 1 , figure 1 It is a schematic flowchart of a method for improving the generalization ability of a convolutional neural network provided by Embodiment 1 of the present invention. As shown in the figure, the method may include the following steps:

[0037] Step S101, read a group of images from the training set, map the group of images into a plurality of image feature vectors, and classify the image feature vectors into a plurality of classes according to the categories of the images.

[0038] In the embodiment of the present invention, it is necessary to construct a convolutional neural network first. The convolutional neural network is a multi-layered neural network, and each layer is a transformation (mapping), usually including convolutional transformation and pooling transformation. Each The transformation process is a kind of processing of the input data of each layer, which is another characteristic expression of the characteristics of the input data....

Embodiment 2

[0062] see figure 2 , figure 2 It is a schematic flowchart of a method for improving the generalization ability of a convolutional neural network provided by Embodiment 2 of the present invention. As shown in the figure, the method may include the following steps:

[0063] Step S201, read a group of images from the training set, map the group of images into multiple image feature vectors, and classify the image feature vectors into multiple classes according to the categories of the images.

[0064] This step is the same as step S101, for details, please refer to the relevant description of step S101, which will not be repeated here.

[0065] Step S202, according to the image feature vectors in each class, calculate the first data feature of all the image feature vectors in the class as the first data feature of the class.

[0066] As a preferred embodiment, the calculation of the first data feature of all image feature vectors in the class according to the image feature v...

Embodiment 3

[0100] see Figure 5 , Figure 5 It is a schematic block diagram of the device for improving the generalization ability of the convolutional neural network provided by the third embodiment of the present invention. For the convenience of description, only the parts related to the embodiment of the present invention are shown.

[0101] The device for improving the generalization ability of the convolutional neural network can be a software unit, a hardware unit or a combination of software and hardware built in terminal equipment (such as mobile phones, tablet computers, notebooks, computers, servers, etc.), or it can be used as an independent The pendant is integrated into the terminal device.

[0102] The device for promoting the generalization ability of the convolutional neural network comprises:

[0103] The classification module 41 is used to read a group of images from the training set, map the group of images into a plurality of image feature vectors, and divide the 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 belongs to the technical field of neural network, and provides a method and apparatus for increasing generalization capability of a convolutional neural network. The method includes the following steps: reading a group of images from a training set, mapping the group of images to a plurality of image character vectors, dividing the plurality of image characteristic vectors into a plurality of classes based on the types of the images; based on the image characteristic vectors of each class, calculating the intra-class loss function of all the characteristic vectors; based on the image characteristic vectors of each class, calculating the inter-class loss function of all the characteristic vectors; based on the intra-class loss function of all the characteristic vectors, using the back propagation algorithm to update the weight of each node of the convolutional neural network; repeating the above mentioned steps until the convolutional neural network converges on the training set or reaches preset repeating times. According to the invention, the method and the apparatus can save all data in long-tailed distribution, makes full usage of abundant inter-class information of tail data, and increases the generalization capability of the convolutional neural network.

Description

technical field [0001] The invention belongs to the technical field of neural networks, and in particular relates to a method and a device for improving the generalization ability of a convolutional neural network. Background technique [0002] Convolutional neural network is a feed-forward neural network. Its artificial neurons can respond to surrounding units within a part of the coverage area, and it has excellent performance for large-scale image processing. Different from the traditional algorithm, the neural units between the adjacent layers of the convolutional neural network are not fully connected, but partially connected, and the weight of the convolution operation for a convolution kernel is shared, thereby reducing the number of parameters, by Multiple convolution and pooling processes achieve the purpose of feature extraction. [0003] With the increasing abundance of large-scale image databases and the continuous improvement of computer computing performance, ...

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): G06N3/08G06K9/62
CPCG06N3/084G06F18/24
Inventor 乔宇张潇
Owner SHENZHEN INST OF ADVANCED TECH CHINESE ACAD OF SCI
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