Visual pattern recognition framework based on the best rank tensor decomposition

Abstract

In this paper a framework for visual patterns recognition of higher dimensionality is discussed. In the training stage, the input prototype patterns are used to construct a multidimensional array—a tensor—whose each dimension corresponds to a different dimension of the input data. This tensor is then decomposed into a lower-dimensional subspace based on the best rank tensor decomposition. Such decomposition allows extraction of the lower-dimensional features which well represent a given training class and exhibit high discriminative properties among different pattern classes. In the testing stage, a pattern is projected onto the computed tensor subspaces and a best fitted class is provided. The method presented in this paper, as well as the software platform, is an extension of our previous work. The conducted experiments on groups of visual patterns show high accuracy and fast response time.