Unlock instant, AI-driven research and patent intelligence for your innovation.
Domain adaptive image classification network training method, image classification method and device
What is Al technical title?
Al technical title is built by PatSnap Al team. It summarizes the technical point description of the patent document.
A classification network and training method technology, applied in neural learning methods, biological neural network models, instruments, etc., can solve problems such as inability to solve the distribution of semantic information at the same level, and achieve the effect of improving classification accuracy
Pending Publication Date: 2022-06-28
ZHEJIANG UNIV
View PDF0 Cites 1 Cited by
Summary
Abstract
Description
Claims
Application Information
AI Technical Summary
This helps you quickly interpret patents by identifying the three key elements:
Problems solved by technology
Method used
Benefits of technology
Problems solved by technology
[0006] The purpose of the embodiment of the present application is to provide a training method, image classification method and device for a domain-adaptive image classification network, so as to solve the technical problem in the related art that the semantic information of the same level is distributed in the output of each layer of the model network
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
Click on the blue label to locate the original text in one second.
Reading with bidirectional positioning of images and text.
Smart Image
Examples
Experimental program
Comparison scheme
Effect test
Embodiment 1
[0069] figure 1 is a flowchart of a training method for a domain-adapted image classification network according to an exemplary embodiment, as shown in figure 1 As shown, the method may include the following steps:
[0070] Step S11: obtaining several pairs of source domain images and target domain images, wherein the categories of each pair of source domain images and target domain images are the same;
[0071] Step S12: extracting a pair of cross-layer features of the source domain image and the target domain image;
[0072] Step S13: using the attention mechanism to calculate the similarity between the cross-layer features of the source domain image and the target domain image;
[0073] Step S14: Calculate the domain alignment generalization loss according to the multi-kernel maximum mean difference of the cross-layer feature and the similarity;
[0074] Step S15: Calculate the classification loss according to the cross-layer features of the source domain image and the t...
Embodiment 2
[0147] Figure 7 is a flow chart of an image classification method according to an exemplary embodiment, such as Figure 7 As shown, the method may include the following steps:
[0148] Step S51: acquiring the target domain image to be classified;
[0149] Specifically, for each target domain image, according to the above steps S21 and S22, first unify the size of the image, then unify the image size and perform normalization to obtain the target domain image x t .
[0150] Step S52: inputting the target domain image into a domain-adapted image classification network, wherein the domain-adapted image classification network is a network obtained by training according to the training method of the domain-adapted image classification network described in Embodiment 1;
[0151] Specifically, the encoding matrix x of the source domain image and the target domain image t Input the feature extractor F for feature extraction to obtain the second target domain feature f t . will ...
Embodiment 3
[0166] Correspondingly, the present application also provides an electronic device, comprising: one or more processors; a memory for storing one or more programs; when the one or more programs are executed by the one or more processors , so that the one or more processors implement the above-mentioned training method or image classification method of a domain-adapted image classification network.
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
Login to View More
Abstract
The invention discloses a training method of a domain adaptive image classification network, and an image classification method and device. The training method of the domain adaptive image classification network comprises the following steps: acquiring a plurality of pairs of source domain images and target domain images; extracting cross-layer features of one pair of source domain image and target domain image; calculating the similarity between the cross-layer features by using an attention mechanism; according to the multi-core maximum mean value difference of the cross-layer features and the similarity, domain alignment generalization loss is calculated; calculating classification loss according to the cross-layer features of the source domain image and the target domain image; according to the generalization loss and the classification loss of domain alignment, performing weighted calculation on the total loss of the domain adaptive image classification network; updating parameters of the domain adaptive image classification network according to the total loss; and executing the step of extracting cross-layer features of one pair of the source domain image and the target domain image to update parameters of the domain adaptive image classification network according to the total loss on the rest of the source domain images and the target domain images until the cross-layer alignment loss is converged.
Description
technical field [0001] The present application relates to the technical field of image classification, and in particular, to a training method, an image classification method, and an apparatus for a domain-adapted image classification network. Background technique [0002] The rapid development of technologies such as machine learning algorithms and deep neural networks has greatly improved the performance of image classification models. When there are enough labeled training samples, and the training samples and test samples satisfy the assumption of independent and identical distribution, the classification model can achieve better results. However, in practical applications, collecting enough labeled training images is often time-consuming, expensive, or even impossible. At the same time, due to various factors, it is impossible to ensure that the training samples always have the same distribution as the test samples, and the difference in data distribution makes it diff...
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
Application Date:The date an application was filed.
Publication Date:The date a patent or application was officially published.
First Publication Date:The earliest publication date of a patent with the same application number.
Issue Date:Publication date of the patent grant document.
PCT Entry Date:The Entry date of PCT National Phase.
Estimated Expiry Date:The statutory expiry date of a patent right according to the Patent Law, and it is the longest term of protection that the patent right can achieve without the termination of the patent right due to other reasons(Term extension factor has been taken into account ).
Invalid Date:Actual expiry date is based on effective date or publication date of legal transaction data of invalid patent.
Login to View More 
Login to View More