Facial expression recognition method and device based on expression intensity label distribution

Abstract

The invention relates to a facial expression analysis method and a device based on expression intensity label distribution, and a computer readable medium. The method comprises the following steps: automatically encoding a face image sequence to generate an expression intensity label; and selecting neutral expressions and non-neutral expressions from the expression sequence as input, training theinput data by adopting a convolutional neural network to obtain an expression recognition model, and predicting the category of the expression image and the intensity of the category according to thejoint expression intensity distribution output by the expression recognition model. According to the method, the training data of expression intensity labeling can be obtained, so that the correctnessand robustness of expression recognition are improved.

Description

technical field [0001] The invention belongs to the technical field of image recognition, and in particular relates to a facial expression recognition method, device and computer-readable medium based on expression intensity label distribution. Background technique [0002] Facial expression recognition is a comprehensive subject involving pattern recognition, image processing, artificial intelligence and other disciplines. The so-called facial expression recognition refers to allowing the computer to extract features from a given expression image and combine the prior knowledge of human beings. , to start the process of learning, reasoning, judgment, and understanding of human emotions. Facial expression recognition is widely used in emotional computing, human-computer interaction, emotional robotics, medical care and other fields, and is a current research hotspot. [0003] Facial expression recognition is mainly composed of two parts: expression recognition and expressio...

AI Technical Summary

Problems solved by technology

[0005] Aiming at at least one problem or improvement demand in the prior art, the present invention provides a facial expression recognition method, device and computer-readable medium based on the distribution of expression intensity labels, which can solve the problem of lack of expression intensity labels in the prior art. Training data, difficult to achieve accurate labeling and low accuracy of facial expression recognition

Images