A Domain Adaptive Subspace Learning Method Based on Interactive Representation

Abstract

A domain adaptive subspace learning method based on interactive representation, which belongs to the technical field of image classification. The present invention establishes two interactive representation models based on low-rank constraints on the source domain and the target domain, which can better coordinate the relationship between data distribution difference; design a distance constraint to model the subspace relationship between the source domain and the target domain, and solve the problem of inconsistency between the two domains; develop a joint framework that uses label-based regression to obtain additional classification discrimination to This facilitates the adaptability and robustness of the model; compared to other methods, the present invention has higher accuracy and more robust performance.

Description

technical field [0001] The invention belongs to the technical field of image classification, and in particular relates to a domain adaptive subspace learning method based on interactive representation. Background technique [0002] In most machine learning methods, the source and target domains are usually considered to have similar distributions. However, in the real world, the source and target domains are affected by light, angle, and background, so there is a problem of domain inconsistency, that is, the source and target domains cannot have similar distributions. However, when classifying images, the inconsistency of the domain will lead to the reduction of the classification accuracy, and the traditional domain adaptation method only considers the local information but not the global information. SUMMARY OF THE INVENTION [0003] The present invention overcomes the deficiencies of the above technologies, and provides a domain adaptive subspace learning method based ...

AI Technical Summary

Problems solved by technology

However, in the real world, the source and target domains are affected by light, angle, background, so the problem of domain inconsistency occurs, that is, the source and target domains cannot have similar distributions

However, when classifying images, domain inconsistency will lead to a decrease in classification accuracy, and traditional domain adaptive methods only consider local information but not global information.

Images