Face expression identification method based on parallel convolutional neural network characteristic graph fusion

Abstract

The invention provides a face expression identification method based on parallel convolutional neural network characteristic graph fusion. The method comprises the steps of designing a convolutional neural network with a parallel structure through simulating vision channels of double eyes of a man, fusing the characteristic graph of the parallel channels after a convolutional pooled layer; on a full connecting layer structure, using sparse full-connection output to one channel, and using dense full-connection output to the other channel, and finally fusing the two outputs and performing classification. After the face expression data are used for performing model training and a relatively high identification rate is realized, a model identification effect is tested by means of a testing sample, and furthermore a relatively high identification accuracy is acquired. The face expression identification method is a new method for emotion analysis and face expression identification.

Description

technical field [0001] The invention belongs to the field of image recognition, in particular to a method for extracting and recognizing facial expression features by using a parallel convolutional neural network. Background technique [0002] Facial expressions contain rich emotional real information, and accurate and efficient recognition of facial expressions is an important research direction in the field of image vision. Facial expression information can be used in many fields such as distance education, auxiliary medical treatment, criminal detection and lie detection. Facial expression recognition technology is a process of classifying and identifying feature information after extracting facial expression features through a specific method. [0003] At present, the commonly used feature extraction methods for facial expression recognition can be divided into methods based on shape models and texture models. Among them, the active appearance model is mainly based on ...

AI Technical Summary

Problems solved by technology

The conventional CNN, DBN, and RNN models have a certain recognition effect on expressions with large differences, but the recognition effect on small expression features is not ideal.

Images