Incremental learning image classification training method under big data scene

A technology of incremental learning and training methods, applied in the field of computer vision, can solve the problems of low performance, implementation, time-consuming and storage of a large amount of training, and achieve the effect of avoiding manual definition of training features and high recognizability

Active Publication Date: 2017-11-17
SOUTH CHINA UNIV OF TECH
View PDF5 Cites 27 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

If the original image data is required for each new category image, it will require a lot of training time and storage, and it cannot be implemented on a machine with lower 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
  • Incremental learning image classification training method under big data scene
  • Incremental learning image classification training method under big data scene
  • Incremental learning image classification training method under big data scene

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0071] The present invention will be further described below in conjunction with specific examples.

[0072] like figure 1 As shown, the image classification training method that can be incrementally learned under the big data scene provided by this embodiment includes the following steps:

[0073] S1, train the initial image classifier, such as figure 2 Shown:

[0074] S1.1. Obtain image data for training and classify according to different image categories;

[0075] This embodiment uses training photos of 101 kinds of flowers downloaded from the Internet, each category contains 1000 photos, and 100 categories of them are selected as initial training data in this step.

[0076] S1.2. Extract features from the image obtained in S1.1 to obtain direct training data; wherein, the features belong to CNN (Convolutional Neural Network, convolutional neural network, specific reference A Krizhevsky, ISutskever, GE Hinton: ImageNet classification with deep convolutional neuralnetw...

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 incremental learning image classification training method under the big data scene. The method comprises steps that step 1, an initial image classifier is trained through utilizing original image data; step 2, if a new class image appears in an application process, incremental training of an initial model is carried out, and the image classifier after update is acquired; and step 3, to-be-classified images are identified through utilizing the trained image classifier, and test image classes are acquired. The method is advantaged in that the convolutional neural network is utilized to carry out image characteristic extraction, manual characteristic definition can be avoided, and the high identification degree is realized; the convolutional neural network is utilized to carry out image identification, memory occupation is small, the calculation speed is fast, incremental leading of the new class image can be carried out, storage of the original training data is not needed, the training time and the storage time can be substantially saved, and the method is especially suitable for big data image classification occasions.

Description

technical field [0001] The invention relates to the field of computer vision, in particular to an image classification training method capable of incremental learning in a big data scene. Background technique [0002] In recent years, the field of computer vision is in a period of rapid development. As an important technology in this field, image classification is designed to allow computers to replace humans to process a large amount of physical information, thereby automatically identifying the subject category in the image. [0003] At present, there are related patents in the field of image classification, such as the automatic image classification method proposed by the patent CN103577475A, which includes the following steps: receiving the image to be classified; reading the feature categories in the feature library; Classify the feature data of the picture; match the extracted feature data with the preset feature data corresponding to the feature category, and merge th...

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 Applications(China)
IPC IPC(8): G06K9/62G06N3/04
CPCG06N3/045G06F18/241G06F18/214
Inventor 郭礼华陈达武
Owner SOUTH CHINA UNIV OF TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap