Face image enhancement method and system based on self-attention network, and medium

Abstract

The invention discloses a face image enhancement method and system based on a self-attention network, and a medium. The method comprises the following steps: constructing an initial generative adversarial model based on the self-attention network; obtaining a first face image through a face database, and training the initial generative adversarial model according to the first face image to obtain a first generative adversarial model; acquiring a second face image of the actual scene, and training the first generative adversarial model according to the second face image to obtain a second generative adversarial model; and obtaining a third face image to be enhanced, and inputting the third face image into the second generative adversarial model to obtain an enhanced fourth face image. According to the invention, the output enhanced image can better reflect the overall details of the face image data, the completeness and accuracy of the face image are guaranteed, and the face image enhancement efficiency is improved. The method can be widely applied to the technical field of image processing.

Description

Technical field [0001] The present invention relates to the field of image processing technology, in particular a face image enhancement method based on a self-attention network, a system and a medium. Background [0002] In the process of building a smart campus, the safe campus is an important part, which includes face detection and recognition, the deployment and control of personnel on campus and the mapping of nodes they pass through on campus, all of which require cameras to capture and then be analyzed and identified by backstage artificial intelligence. However, in actual engineering projects, many times due to the movement or obstruction of objects, or poor capture angles or dim light, or pure samples are too small, often can not capture enough high-quality images to complete the efficient training of the recognition model, which requires a method of data enhancement based on existing image data. [0003] Existing common data enhancement schemes: In the business scenari...

AI Technical Summary

Problems solved by technology

However, the generative confrontation model based on the convolutional neural network has only a local receptive field, and the deep details of the image will be lost with the training of the model, which affects the integrity and accuracy of the image.

Images