Eureka AIR delivers breakthrough ideas for toughest innovation challenges, trusted by R&D personnel around the world.

Improved CNN-based facial expression recognition method

A facial expression recognition and facial expression technology, applied in the field of facial expression recognition, can solve the problem of small number of originals, and achieve the effect of reducing model parameters, increasing width, and reducing network parameters

Inactive Publication Date: 2018-06-01
CHONGQING UNIV OF POSTS & TELECOMM
View PDF8 Cites 62 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Existing facial expression databases such as CK+, Japanese female facial expression (JAFFE) database, etc., the collected facial expression image samples are collected in an ideal environment and the original number is relatively small, and the training model is more likely to achieve overfitting
Therefore, although the current deep convolutional neural network model trained based on CK+ facial expression database has a relatively high accuracy rate (95%) in the test results, it does not mean that it is competent for the facial expression recognition task in the real environment.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Improved CNN-based facial expression recognition method
  • Improved CNN-based facial expression recognition method
  • Improved CNN-based facial expression recognition method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0036] The technical solutions in the embodiments of the present invention will be described clearly and in detail below with reference to the drawings in the embodiments of the present invention. The described embodiments are only some of the embodiments of the invention.

[0037] The technical scheme that the present invention solves the problems of the technologies described above is:

[0038] like Figure 7 As shown, the present invention provides a kind of facial expression recognition method based on improved CNN, it is characterized in that, comprises the following steps:

[0039] S1: Use a face detection and alignment algorithm JDA algorithm that integrates face detection and alignment functions to obtain facial expression images from video streams;

[0040] S2: Correct the facial posture in the real environment, remove the background information irrelevant to the expression information, and adopt scale normalization. In this embodiment, the specific method is:

[0...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention provides an improved CNN-based facial expression recognition method, and relates to the field of image classification and identification. The improved CNN-based facial expression recognition method comprises the following steps: s1, acquiring a facial expression image from a video stream by using a face detection alignment algorithm JDA algorithm integrating the face detection and alignment functions; s2, correcting the human face posture in a real environment by using the face according to the facial expression image obtained in the step s1, removing the background information irrelevant to the expression information and adopting the scale normalization; s3, training the convolutional neural network model to obtain and store an optimal network parameter before extracting feature of the normalized facial expression image obtained in the step s2; s4 loading a CNN model and the optimal network parameters obtained by s3 for the optimal network parameters obtained in the steps3, and performing feature extraction on the normalized facial expression images obtained in the step s2; s5, classifying and recognizing the facial expression features obtained in the step s4 by using an SVM classifier. The method has high robustness and good generalization performance.

Description

technical field [0001] The invention belongs to the field of image processing and pattern recognition, in particular to a facial expression recognition method based on an improved CNN. Background technique [0002] Human face Facial expression is one of the main non-verbal communication ways to express emotion and information, and it occupies 55% of the information. It can be seen that facial expressions, as an information carrier, play an important role in people's daily communication. Ekman et al. identified six human facial expressions (ie, anger, disgust, fear, happiness, sadness, and surprise) as universal basic human emotion expressions. Facial expression recognition has become a research hotspot in the field of computer vision for many years. Among them, the facial expression recognition system has a wide range of applications, such as human-computer interaction, developmental psychology, biology, medicine, information accessibility, and smart cities. It has a wide ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/00G06K9/62G06N3/04G06N3/08G06T3/40G06T5/00G06T5/40
CPCG06N3/08G06T3/4007G06T5/40G06T2207/10016G06T2207/20081G06T2207/20084G06T2207/30201G06V40/165G06V40/174G06V40/168G06N3/048G06N3/045G06F18/2411G06F18/24323G06T5/80
Inventor 张毅丁剑飞罗元
Owner CHONGQING UNIV OF POSTS & TELECOMM
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Eureka Blog
Learn More
PatSnap group products