An Edge based 3D Facial Expressions Detection System

A.N.V.Susmitha, Mr. K.Srinivas

Abstract Facial expression identification is a process of generating new facial features from a given face and still retain the facial characteristics of the face. Traditioanl 2D based facial expression evaluation encounter difficulties with large variations in facial features in terms of illumination, pose and expression evaluation. By using 3D model the information of human faces head pose and lighting changes are used to find the large set of expressions. In most of the 3D models,face data are projected on to 2D and finds the expression in the 3D. In this model, a novel method is used to find the 3D expression using edge based model for complex pose datasets. Experimental results show that proposed model has high accuracy compare to traditional 3D expression evaluation models. Keywords-3D faces, Ridges, Edge based facial evaluations.

Title:	An Edge based 3D Facial Expressions Detection System
Author Name:	A.N.V.Susmitha, Mr. K.Srinivas
Abstract:	Abstract Facial expression identification is a process of generating new facial features from a given face and still retain the facial characteristics of the face. Traditioanl 2D based facial expression evaluation encounter difficulties with large variations in facial features in terms of illumination, pose and expression evaluation. By using 3D model the information of human faces head pose and lighting changes are used to find the large set of expressions. In most of the 3D models,face data are projected on to 2D and finds the expression in the 3D. In this model, a novel method is used to find the 3D expression using edge based model for complex pose datasets. Experimental results show that proposed model has high accuracy compare to traditional 3D expression evaluation models. Keywords-3D faces, Ridges, Edge based facial evaluations.
Cite this article:	A.N.V.Susmitha, Mr. K.Srinivas , " An Edge based 3D Facial Expressions Detection System " , International Journal of Emerging Trends & Technology in Computer Science (IJETTCS), Volume 4, Issue 6, November - December 2015 , pp. 039-043 , ISSN 2278-6856.
Full Text [PDF] Home