Face emotion recognition method and device based on deep learning, equipment and medium

Abstract

The invention discloses a face emotion recognition method and device based on deep learning, equipment and a medium. The method comprises the following steps of: inputting a to-be-identified image into a preset deep learning model, performing channel data extraction on the to-be-identified image by using an input layer in the deep learning model; obtaining face image data of the to-be-identified image; carrying out depth convolution calculation on the face image data by using the first convolution core of the convolution layer to obtain feature data, carrying out point convolution calculationon the feature data by using the second convolution core of the convolution layer to obtain face features of the to-be-identified image, and finally, carrying out classification regression on the facefeatures by using a full connection layer to obtain the emotional state of the to-be-identified image. According to the embodiment of the invention, the standard convolution operation in the traditional deep learning algorithm is decomposed into the calculation of one deep convolution and the calculation of one point convolution, so that the calculation amount of the deep learning model is reduced, and the operation rate of the emotion recognition model is improved.

Description

technical field [0001] The present invention relates to the technical field of biometrics, in particular to a method, device, equipment and medium for facial emotion recognition based on deep learning. Background technique [0002] Deep learning is currently a relatively popular application field in the field of artificial intelligence. It has better results in speech recognition, image recognition and natural language processing. The recognition of facial emotions is to study how to make the computer learn from static images or video sequences. The technology of obtaining and distinguishing facial expressions is of great significance to the research of human-computer interaction and emotional computing. At present, deep learning algorithms are generally used to train emotion recognition models, such as convolutional neural networks and recurrent neural networks. Deep learning algorithm, through deep learning of face sample pictures, builds an emotion recognition model to de...

AI Technical Summary

Problems solved by technology

[0004] Embodiments of the present invention provide a face recognition method, device, device and medium based on deep learning to solve the problem of large amount of computation for emotion recognition models constructed by deep learning algorithms

Images