Glasses removing method for fine-grained face recognition

Abstract

The invention belongs to the technical field of face recognition, and particularly relates to a glasses removal method for fine-grained face recognition, which comprises the following steps: segmenting an initial glasses-worn face image into three image blocks, respectively identifying the three image blocks by Part1, Part2 and Part3, wherein the Part2 comprises a complete glasses part; establishing an ERCNN (Eyeglass Removing Deep Convolutional Neural Network) model; taking the Part2 as the input of the convolution layer of the ERCNN network model; performing feature selection and maximum element operation through an MFM unit in the network, and reconstructing the Part2 by utilizing operations of deconvolution, average pooling and weighted summation of elements one by one, thereby obtaining a new image block Part2_new after the glasses are removed; and combining the output Part2_new with the original Part1 and the original Part3 to obtain a complete face image of which the glasses areremoved.

Description

technical field [0001] The invention belongs to the technical field of face recognition, and in particular relates to a method for removing glasses for fine-grained face recognition. Background technique [0002] Face recognition is one of the most active research directions in the field of pattern recognition and computer vision. As a common face occluder, glasses greatly affect the accuracy of face recognition, especially for face recognition with fine-grained features such as similar faces. Currently, there are methods based on PCA or deep learning for removing glasses from face images. Among them, PCA is a more commonly used data analysis method. Its main idea is to calculate the principal component components, that is, the transformation matrix, according to statistical principles, so as to reconstruct the original vector. Although the PCA method can remove glasses from face images wearing glasses, it is susceptible to noise interference, and the removal effect is not...

AI Technical Summary

Problems solved by technology

Compared with the technical solution of this proposal, this technical solution has the following disadvantages: 1. "Application No. 201711361308.9" adopts a symmetrical structure of N convolutional layers and N deconvolutional layers, and the number of network layers is large, so it is not easy to implement Training, and the trained model parameters and calculations are large

2. The glasses removal method proposed by "Application No. 201711361308.9" only applies it to general face recognition methods, and does not apply the reconstructed glasses-free face image to fine-grained face recognition, and does not define the effect on the reconstructed image. Quality Evaluation Criteria

Images