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

Novel One-Shot learning method based on SimpleShot

A new method and base class technology, applied in the field of target recognition technology for small-sample learning, can solve problems such as over-fitting, achieve the effects of improving accuracy, alleviating over-fitting problems, and improving classification accuracy

Pending Publication Date: 2020-07-28
WUHAN UNIV
View PDF3 Cites 7 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] In order to overcome the over-fitting problem that is easy to occur in small-sample learning and improve the classification accuracy of existing small-sample image classification methods, the present invention proposes a new One-Shot learning method based on SimpleShot

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
  • Novel One-Shot learning method based on SimpleShot
  • Novel One-Shot learning method based on SimpleShot
  • Novel One-Shot learning method based on SimpleShot

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0029] The technical scheme of the present invention will be further described below in conjunction with the accompanying drawings.

[0030] Such as figure 1 As shown, the present invention provides a small sample learning method based on SimpleShot, mainly for one-shot learning, including the following steps:

[0031] S100: Suppose we get a data set D formed by a base class base ={(I 1 ,y 1 ),..., (I N ,y N ))}, which contains N labeled images from the base class A, namely y n ∈{1,...,A},I N Indicates the Nth image, y N For the label of the Nth image, use EfficientNets as the feature extraction network f θ After training on this base class, this trained model is called feature space;

[0032] The implementation of further step S100 is as follows:

[0033] Using the base class to train the feature network is carried out in the normal steps of the usual image classification, as follows:

[0034] Suppose we get a data set D base ={(I 1 ,y 1 ),..., (I N ,y N ))}, ...

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 a new One-Shot learning method based on Simpleshot. A feature extraction network Effect Nets with optimal performance in the field of image classification at present is used asa feature extraction network; training is carried out on the base class to obtain a feature space, then small samples of a new class are input, feature transformation is carried out on the image in the feature space, namely, the regularization technology is used for relieving the overfitting problem, and finally nearest neighbor classification is carried out.

Description

technical field [0001] The present invention relates to the SimpleShot algorithm, a target recognition technology that re-studies nearest neighbor classification for small-sample learning. This technology is one of the important technologies in the field of small sample learning. By re-examining the nearest neighbor classification and combining some novel ideas, it has advanced accuracy in the field of small sample learning. Background technique [0002] In recent years, with the development of deep learning, more and more deep learning models have achieved very good results in the field of computer vision, especially in the field of image classification, and have achieved the most advanced performance. The recognition accuracy can even reach A state rivaling human vision. However, this phenomenon is due to the fact that to achieve extremely powerful performance, it is necessary to rely on a large amount of labeled data to train the model, and the model requires a network w...

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): G06K9/62G06K9/46G06N3/04G06N3/08
CPCG06N3/08G06V10/462G06N3/045G06F18/24147
Inventor 于秋则张杰豪王欢倪达文
Owner WUHAN 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