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

Single sample image segmentation method and system with independent training data set

A training data set and image segmentation technology, which is applied in image analysis, image enhancement, image data processing, etc., can solve the problems of not considering the training data set, performance degradation, time-consuming and labor-intensive, etc., and achieve a large decrease in segmentation performance and good segmentation Performance, the effect of reducing distribution differences

Pending Publication Date: 2022-01-11
SOUTH CHINA UNIV OF TECH
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

This is undoubtedly time-consuming
Since the current models do not consider the difference between the training dataset and the target dataset, these models suffer from significant performance degradation when predicting the target dataset

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
  • Single sample image segmentation method and system with independent training data set
  • Single sample image segmentation method and system with independent training data set
  • Single sample image segmentation method and system with independent training data set

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0061] Such as figure 1 As shown, a single-sample image segmentation method independent of the training data set, including the following steps:

[0062] S1 establishes training data and test data. The two data come from different data sets. Both the training data and the test data are divided into a support set and a query set. The two data sets have distribution differences.

[0063] In this embodiment, two data sets are used as model test data sets, PASCAL VOC 2012, and 2D CT lung lobe data set. The PASCAL VOC 2012 dataset contains 1464 training images and segmentation labels for 20 categories. CT lobe data were obtained from the Lung Nodule Analysis (LUNA) competition. The dataset contains 534 2D CT images and corresponding masks. As shown in Figure 2(a) and Figure 2(b) the training data comes from PASCAL VOC 2012) and Figure 2(c) (the test data comes from the CT lung lobe data set), you can intuitively feel the difference between the two data sets

[0064] S2 builds a...

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 a single sample image segmentation method and system with independent training data set, and the method comprises the steps: building training data and test data which are from different data sets, and dividing the training data and the test data into a support set and a query set; constructing a segmentation branch network model and a distribution alignment branch network model; training a segmentation branch network model and a distribution alignment branch network model; and utilizing the trained segmentation branch network model to predict the category of test data. The deep network trained by the method can solve the problem of large distribution difference between a training data set and a test data set in single-sample image segmentation, and the segmentation performance is further improved.

Description

technical field [0001] The invention relates to the field of semantic segmentation, in particular to a single-sample image segmentation method and system with independent training data sets. Background technique [0002] The traditional semantic segmentation methods using deep learning all adopt the structure of fully convolutional network, and have been able to achieve better segmentation results, but the model based on the fully convolutional structure requires a large amount of labeled data to obtain satisfactory results. Effect. However, in real life, labeled semantic segmentation data is very expensive, which requires human labor to carefully annotate pixel-level labels. Also, for some fields, such as medicine, the number of available samples is very scarce due to privacy issues and rare diseases. Although there have been some works to address these issues, such as weakly supervised semantic segmentation, a large amount of relevant weakly labeled data is still require...

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): G06T7/10G06V10/764G06V10/74G06V10/774G06N3/04G06N3/08
CPCG06T7/10G06N3/08G06T2207/10081G06T2207/30061G06N3/045G06F18/22G06F18/214G06F18/24
Inventor 陈琼杨咏冼进
Owner SOUTH CHINA UNIV OF TECH
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