Training method for two-dimensional face recognition model based on deep convolutional neural network

Abstract

The invention provides a training method for a two-dimensional face recognition model based on a deep convolutional neural network. The method comprises the steps that face images are collected and converted into grayscale images, the grayscale images are input into a face convolutional neural network model for training, a triad most difficult to distinguish is constructed, the face recognition convolutional neural network model obtained after last round of training is utilized to perform a next round of triad selection and training by means of iterative optimization till iteration is converged, and a final face recognition convolutional neural network model used for recognition is obtained. Compared with the prior art, large-scale face images with different expressions and postures are effectively utilized, the effective method capable of being used for training the two-dimensional face recognition model is proposed, precise feature expression can be effectively learned by means of iterative optimization, face comparison precision is improved, and an optimal model on a sample set can be obtained as long as sufficient sample images and iterations are available.

Description

technical field [0001] The invention relates to the field of biological feature recognition in pattern recognition, and mainly relates to a training method for a two-dimensional face recognition model based on a deep convolutional neural network. Background technique [0002] Face recognition is mainly used for identification, especially in recent years, with the rapid progress of computer technology, image processing technology, pattern recognition technology, etc., a new biometric recognition technology has emerged. Because it can be widely used in many fields such as security verification, video surveillance, access control, etc., with fast recognition speed and high recognition rate, it has become the main development direction in the field of identification technology research. [0003] The current mainstream face recognition acquires face images on the basis of cooperation, and applies classification algorithms for face recognition. There are mainly the following type...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to overcome the deficiencies of the prior art and provide a training method for a two-dimensional face recognition model based on a deep convolutional neural network to solve the problem of face Technical problems such as low recognition rate and poor reliability

Images