Face image processing method and device and storage medium

Abstract

The embodiment of the invention discloses a face image processing method and device and a storage medium, and belongs to the technical field of image processing. The method comprises the steps of obtaining a first sample face image and a corresponding second sample face image; based on the trained first feature extraction model, performing feature extraction on the first sample face image to obtain a first sample identity feature of the first sample face image; based on a current second feature extraction model, performing feature extraction on the second sample face image to obtain a second sample identity feature of the second sample face image; and adjusting model parameters of the second feature extraction model according to an error between the first sample identity feature and the second sample identity feature. According to the method, the second feature extraction model can learn the relation between the face image and the corresponding identity feature, a more accurate featureextraction model can be trained, the identity feature extracted by the second feature extraction model is more accurate, and the accuracy is improved.

Description

technical field [0001] Embodiments of the present invention relate to the technical field of image processing, and in particular, to a face image processing method, device, and storage medium. Background technique [0002] With the development of image processing technology and the widespread popularization of intelligent equipment, face images will be processed in various scenarios such as face detection, face recognition, and facial expression analysis. However, if the resolution of the collected face image is low and the image quality is poor, the processing effect of the face image will be poor. Therefore, how to process low-resolution face images has become an urgent problem to be solved. [0003] In order to improve image quality, a high-resolution face image is usually obtained from a low-resolution face image. see figure 1 , the related technology will first train the residual network model, the residual network model learns the difference between the low-resoluti...

AI Technical Summary

Problems solved by technology

[0004] When the above solution obtains the high-resolution face image corresponding to the low-resolution face image, it does not consider the identity features in the face image that can identify the identity of the face, which may affect the subsequent face image processing effect

Images