Four-dimensional Facial Real Expression Recognition Based On Expressional Motion Rate

Facial expressions on the face contain very rich human and psychological information,which can be used as a basis for judging and measuring human emotions,inner activities,emotions,and physical health.In some cases,we can even communicate only by facial expressions without language.There are two kinds of expressions: real and unreal.Real expressions refer to natural expressions without intervention,usually including microexpressions and uncontrolled general expressions.Microexpressions can be understood as the complete expression of real expressions that human beings try to suppress or conceal for a certain purpose.In the short period from the generation of expressions to the blocked expression,microexpressions are generated.Compared with general expressions,microexpressions move faster.Non-real expression refers to the result of the intervention of natural expression.Previous people have classified a variety of non-real expressions,which are deliberate,posed,false,social,polite and masking.In recent years,real facial expression recognition has become more and more important in public service,medical field,psychological analysis and other fields.However,neither the design of basic real facial expression database nor the core recognition methods have been studied from the movement characteristics of facial expressions.This limitation restricts the further development of real facial expression recognition.This paper is based on these problems.Research and exploration.In view of the above problems,this paper firstly regards expression as a kind of movement,and studies it from the perspective of expression authenticity,unifies and integrates general expression and micro-expression,and then puts forward the definition of real expression.Then,a four-dimensional real facial expression recognition model is proposed to describe the trajectory of the expression,and to measure the authenticity of different expressions and expressions by the expression movement rate.An algorithm for calculating the expression movement rate and a method for building a four-dimensional real facial expression recognition model are designed.Finally,two groups of experiments are designed to verify the real expression recognition method.The experimental results show that the four-dimensional face real expression recognition model based on the speed of expression movement can not only better describe and process the trajectory and data generated by expression movement,but also improve the speed and accuracy of real expression recognition.