| | |
Summary: Compact Representation of Multidimensional Data Using Tensor Rank-One
Decomposition
Hongcheng Wang, Narendra Ahuja
Beckman Institute, University of Illinois at Urbana-Champaign, USA
{wanghc,ahuja }@vision.ai.uiuc.edu
Abstract
This paper presents a new approach for representing
multidimensional data by a compact number of bases. We
consider the multidimensional data as tensors instead of
matrices or vectors, and propose a Tensor Rank-One De-
composition (TROD) algorithm by decomposing Nth-order
data into a collection of rank-1 tensors based on multilin-
ear algebra. By applying this algorithm to image sequence
compression, we obtain much higher quality images with
the same compression ratio as Principle Component Analy-
sis (PCA). Experiments with gray-level and color video se-
quences are used to illustrate the validity of this approach.
1. Introduction
In computer vision and graphics, we often encounter
multidimensional data, such as images, video, range data
|
