Human face shielding degree recognition method and device, electronic equipment and readable storage medium

Abstract

The invention provides a human face shielding degree recognition method and device, electronic equipment and a readable storage medium. The method comprises the steps of: obtaining a to-be-recognizedimage; inputting the to-be-recognized image into a neural network model, and obtaining the face shielding degree of the to-be-recognized image based on the output result of the trained neural networkmodel, wherein the neural network model is obtained by training a preset number of face occlusion data, and the face occlusion data is obtained by the following steps: obtaining a preset number of original images; and after aligning the human faces in each original image, shielding the aligned human faces by using the shielding object image, and obtaining a preset number of human face shielding data based on each aligned human face and each shielded human face. According to the scheme, the acquisition of the face shielding data is simpler and more efficient, so that the neural network model for face shielding degree recognition can be obtained through more efficient training, and the neural network model is more favorably applied to face shielding degree recognition.

Description

technical field [0001] The present disclosure relates to the field of computer technology, and in particular, the present disclosure relates to a face occlusion degree recognition method, device, electronic equipment and readable storage medium. Background technique [0002] In the video live broadcast industry, live broadcast platforms often require users to upload avatar pictures, and the platform can identify users' identities through the faces in the avatar pictures. However, the avatar pictures uploaded by users often contain covering objects such as hats, glasses, masks, hair, etc. These covering objects will block some areas of the user's face. If the platform cannot identify the user's identity, the platform will determine that the user's avatar picture is an unqualified avatar picture. [0003] At present, many platforms use neural networks to identify the occlusion degree of the face in the avatar picture uploaded by the user. If the occlusion degree is greater th...

AI Technical Summary

Problems solved by technology

Before using the neural network to identify the degree of occlusion of the face, a large amount of face occlusion data (that is, images marked with the degree of occlusion of the face) is required as training samples to train the neural network. It consumes manpower and material resources, and the labeling efficiency is low, which is not conducive to the use of neural networks to identify the degree of face occlusion

Images