Simple face recognition online learning method based on deep learning model

Abstract

The invention is applicable to the technical field of computer vision, and provides a simple face recognition online learning method based on a deep learning model. Face feature vectors and identity label information are stored by using files capable of adding data types, such as hdf5 and the like; the human face feature vectors and the corresponding identity tags are added to the local human facefeature vector library with the same index, so that members are added and deleted more simply and conveniently, the trouble of re-collecting samples and training again in a traditional method is solved, and the method is simple to operate, easy to master, remarkable in effect. The updating of the local face feature vector library does not need to carry out face feature extraction on all local members again, only face feature extraction is carried out on newly added personnel, and the face feature vectors and the identity label information of the newly added members are added to the local facefeature vector library. And the updating of the local face feature vector library is greatly facilitated.

Description

technical field [0001] The invention belongs to the technical field of computer vision, and in particular relates to a simple face recognition online learning method based on a deep learning model. Background technique [0002] Facial recognition refers to biometric technology that uniquely identifies or verifies a person by comparing and analyzing patterns based on their facial contours. After decades of development, face recognition technology has become more and more mature, and its application scope has become more and more extensive. However, because the face is different from fingerprints, irises and other biological characteristics, its changes are diverse, and it is also affected by illumination, posture, and occlusion. The accuracy of face recognition is difficult to guarantee. In order to increase the robustness of the algorithm, it is necessary to update the model. But for traditional methods, model update means resampling and then training the model, which is v...

AI Technical Summary

Problems solved by technology

[0004] In view of the above problems, the object of the present invention is to provide a simple face recognition online learning method based on a deep learning model, which aims to solve the technical problem that the existing face recognition technology needs to repeatedly train the model when an error case occurs

Images