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

Twin classifier certainty maximization method for cross-domain complex vision task

A classifier and deterministic technology, applied in the field of transfer learning, can solve the problems of single adaptation scene and insufficient discriminability, and achieve the effect of improving model performance, ensuring discriminability and predicting diversity.

Pending Publication Date: 2021-05-14
BEIJING INSTITUTE OF TECHNOLOGYGY
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0009] In order to solve the problem that the cross-domain visual task adapts to a single scene and the discriminability of the feature representation is insufficient, the present invention proposes a twin classifier certainty maximization method for cross-domain complex visual tasks

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
  • Twin classifier certainty maximization method for cross-domain complex vision task
  • Twin classifier certainty maximization method for cross-domain complex vision task
  • Twin classifier certainty maximization method for cross-domain complex vision task

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0040] In order to make the purpose, technical solutions and advantages of the present invention clearer, specific examples of the method of the present invention will be further described in detail.

[0041] For ease of understanding, in this example, include a source domain with the label where: n s is the sample size, is the i-th sample in the source domain, is the corresponding label, and an unlabeled target domain where: n t is the number of samples in the target domain, is the i-th sample in the target domain; the goal of the method of the present invention is to migrate the deep neural network model trained on the source domain samples to the target domain, and enable it to learn a good transferable Feature representation of sex and discriminability, so as to achieve good performance of the model on the target domain, that is, φ:X t →Y t ; The model framework of the method of the present invention comprises a feature generator G and two twin classifiers C ...

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 relates to a twin classifier certainty maximization method for a cross-domain complex visual task; the method comprises the following steps: 1, constructing a neural network frame, and initializing the neural network frame; 2, inputting the source domain sample into a feature generator G to obtain a corresponding feature representation; and 3, under the supervision of the label information of the source domain sample, calculating an empirical risk error between a model prediction output p and a real label y on the source domain sample by using a standard cross entropy loss function. The method has the advantages that a novel classifier deterministic difference measurement CDD is designed, wherein the difference of the classifiers is measured by means of category correlation between target prediction of the twin classifiers, and implicit constraints are applied to the identifiability of target features at the same time.

Description

technical field [0001] The invention relates to an unsupervised field adaptive matching method in the field of transfer learning, specifically, relates to a twin classifier certainty maximization method for complex visual tasks of image classification, semantic segmentation and target detection. Background technique [0002] With the rapid development of information technology and the substantial growth of data scale, machine learning has made great progress both in theory and in practical applications, especially the Deep Neural Network (DNN) proposed in recent years, It has been successfully applied in many fields including computer vision, natural language processing, medical diagnosis, etc., and has achieved great breakthroughs. However, it is worth noting that the great success of deep learning relies heavily on large-scale labeled data. However, in many practical application scenarios, the acquisition of labeled data requires expensive time and labor costs, which lead...

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
CPCG06N3/045G06F18/2415G06F18/241
Inventor 李爽刘驰吕芳蕊
Owner BEIJING INSTITUTE OF TECHNOLOGYGY
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