Paper Info
Title
Tensor-Based Active Appearance Model
Abstract
The active appearance model (AAM) is a well-known model that can represent a nonrigid object effectively. However, the fitting result is often unsatisfactory when the input image has pose, expression, and illumination variations. To overcome this problem, we propose a tensor-based AAM which consists of two kinds of tensors: image tensor and model tensor. The image tensor is used to estimate the image variation such as the pose, the expression, and the illumination by finding the basis subtensor with minimal reconstruction error. The model tensor is used to generate the specific AAM basis vectors by indexing the model tensor in terms of the estimated image variations. Experimental results show that the proposed tensor-based AAM reduces the fitting error of the conventional AAM by about four pixels.
Year
DOI
Venue
2008
10.1109/LSP.2008.2001116
IEEE Signal Process. Lett.
Keywords
Field
DocType
active appearance model (aam),tensor,active appearance model,multilinear analysis,tensor-based aam,nonrigid object,tensor-based active appearance model,fitting error,image variation,image reconstruction,model tensor,minimal reconstruction error,image tensor,tensors,face,shape,face recognition,indexation,computational modeling,biometrics,knowledge engineering,tensile stress,lighting,indexing,face detection,fitting
Iterative reconstruction,Computer vision,Pattern recognition,Tensor,Search engine indexing,Stress (mechanics),Reconstruction error,Active appearance model,Artificial intelligence,Pixel,Basis (linear algebra),Mathematics
Journal
Volume
ISSN
Citations 
15
1070-9908
3
PageRank 
References 
Authors
0.45
8
2
Name
Order
Citations
PageRank
Hyung-Soo Lee113213.10
Daijin Kim21882126.85