Reconstruction of 3D faces using face space coefficient, texture space and shape space

Authors:
Sheng Hung Chung;Ean Teng Khor
Affiliations:
School of Science and Technology, Wawasan Open University, Penang, Malaysia;School of Science and Technology, Wawasan Open University, Penang, Malaysia
Venue:
IVIC'11 Proceedings of the Second international conference on Visual informatics: sustaining research and innovations - Volume Part II
Year:
2011

Citing 6
Cited 0

Closed-Form Solutions for Physically Based Shape Modeling and Recognition

IEEE Transactions on Pattern Analysis and Machine Intelligence
Shape and Nonrigid Motion Estimation Through Physics-Based Synthesis

IEEE Transactions on Pattern Analysis and Machine Intelligence
Modal Matching for Correspondence and Recognition

IEEE Transactions on Pattern Analysis and Machine Intelligence
3D Facial Expression Recognition Based on Primitive Surface Feature Distribution

CVPR '06 Proceedings of the 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition - Volume 2
PCA-Based 3D Face Photography

SIBGRAPI '08 Proceedings of the 2008 XXI Brazilian Symposium on Computer Graphics and Image Processing
Automatic 3D reconstruction for face recognition

FGR' 04 Proceedings of the Sixth IEEE international conference on Automatic face and gesture recognition

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper presents a 3D face reconstruction face model by using PCA-based reconstruction model in synthesizing faces of individual. A 3D face reconstruction model is derived by transforming the shape and texture of the training sets into a vector space representation. In this paper, a reconstruction of face model is adapted from 3 Dimensional Face Space (3DFS) with the knowledge of the shape and texture of faces. Faces statistics produced by sampling from Face Space is computed by Principle Component Analysis (PCA) of 100 exemplar 3D faces. 3D face space is formed by two distinctive subspaces: the 3D shape space and 3D texture space which consist of 79- dimensional (79 shape and texture coefficient). The first shape space shows the impact of the shape and texture dimensions and the second texture space shows the influence of the shape and texture dimensions. A vertex is a point where two edges of a 2D polygon or two or more vertices of a 3D polyhedron meet. Face Space Coefficient (FSC) is computed as the input for training in generating novel 3D faces as Wavefront Object files (OBJ). The output 3D face space is computed with the aid of material file (.mtl) and texture file (.jpg, .rgb) and viewed by OBJ viewer to achieve good 3D representation using this approach.