Higher order support tensor regression for head pose estimation

Weiwei Guo, Irene Kotsia, Ioannis PatrasResearch 2011

Abstract

In this paper, we exploit the advantages of tensor representa-tions and propose a Supervised Multilinear Learning Model for regression. The model is based on the Canonical (CAN-DECOMP)/Parallel Factors (PARAFAC) decomposition of ten-sors of multiple modes and allows the simultaneous projec-tion of an input tensor to more than one discriminative direc-tions along each mode. These projection weights are obtained by optimizing a ϵ-insensitive loss functions which leads to generalized Support Tensor Regression (STR). The methods are validated on the problems of head pose estimation using real data from publicly available databases. 1.
BibTeX
@inproceedings{guo2011higher,
  author = {Weiwei Guo and Irene Kotsia and Ioannis Patras},
  title = {{Higher order support tensor regression for head pose estimation}},
  booktitle = {Research},
  year = {2011},
}

Loading related content...

Figure