Face attribute recognition method and device

Abstract

The embodiment of the invention provides a face attribute recognition method and device, and the method comprises the steps: carrying out face key point detection of a to-be-recognized target image, obtaining face key point information in the target image, and enabling the face key point information to serve as first face key point information; performing feature extraction on the first face key point information to obtain a first face key point feature, and performing feature extraction on the target image to obtain a first original image feature; performing feature fusion on the first face key point feature and the first original image feature to obtain a target feature; and determining a face attribute category in the target image based on the target feature and a pre-trained classification model. Based on the above processing, the precision of face attribute recognition can be improved.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to a face attribute recognition method and device. Background technique [0002] With the rapid development of AI (Artificial Intelligence, artificial intelligence) technology, the face attributes in the image can be identified to determine the face attribute category in the image, and the face attribute category can include the attributes of the facial features of the face in the image Category, for example, can determine whether the hairstyle of the person in the image is parted in the middle or on the side, whether the face shape of the person is round or square, and so on. [0003] In related technologies, machine learning algorithms can generally be used to identify human face attributes. In one way, the convolutional neural network model can be used to extract the features of the face in the image, and according to the extracted face features and classificati...

AI Technical Summary

Problems solved by technology

[0004] However, in related technologies, limited by the accuracy of the convolutional neural network model, the extracted features may not accurately reflect the local features of the face in the image, for example, the shape of the character's beard, the shape of the character's eyebrows and other local features. , which in turn leads to lower accuracy of face attribute recognition

Images