Multi-source domain adaptive face recognition method

Abstract

The invention belongs to the technical field of artificial intelligence, and the domain adaptive method of the invention also belongs to an important branch of the migration learning field, and discloses a multi-source domain adaptive face recognition method. The paper presents a method of learning the common subspace of multiple source domains and a single target domain, which is applied to facerecognition problem. It solves the problem that the classifier which learns from the source domain has poor recognition effect in the target domain when the data of the source domain and the target domain are distributed differently and the data of the target domain does not have or has a small number of labels. By learning samples from multiple source domains and single target domains, a common subspace is obtained. By transforming the source domain into the linear representation of the target domain in the common subspace, the target source domain data is casted into the high dimensional space to construct the linear optimal hyperplane in the high dimensional space, and a kernel classifier is learned. Then the test samples are identified and classified, and the beneficial results are obtained.

Description

technical field [0001] The invention belongs to the technical field of artificial intelligence and relates to a face recognition method, in particular to a domain adaptive face recognition method. Background technique [0002] Now some traditional machine learning methods such as Fisher linear discriminant analysis, support vector machine, BP network, etc. have certain applications in the field of face recognition. The general steps of these methods are as follows: first, train the labeled source domain samples to learn a better classifier, and then test the target domain samples with the same distribution as the source domain samples to obtain the face recognition results. However, in real applications, the training data and test data often do not satisfy the condition of the same distribution. In the training process of the existing methods, the training samples often have good performance, but in the test, due to the different If the test sample data is distributed, the ...

AI Technical Summary

Problems solved by technology

However, in real applications, the training data and test data often do not satisfy the condition of the same distribution. In the training process of the existing methods, the training samples often have good performance, but in the test, due to the different Distributed test sample data, the classifier will degenerate, and the effect will be greatly reduced

A better solution is to learn samples with the same distribution as the test data, but for the training samples, the source doma...

Images