In this work we explore how the architecture proposed in [8], which expresses the processing steps of the classical Fisher vector pipeline approaches, i.e. dimensionality reduction by principal component analysis (PCA) projection, Gaussian mixture model (GMM) and Fisher vector descriptor extraction as network layers, can be modified into a hybrid network that combines the benefits of both unsup...

Abstract. Differential algebraic equations have wide applications in the field of engineering and science where the mathematical models form the descriptor systems. Analysis and modeling of the solutions of such systems need to handle with the different equations related to the systems. A time-domain discretization, for example, finite difference, finite volume, etc., may lead to DAEs of descri...

In this talk, matrix inequality characterizations for the positive realness of descriptor systems are introduced. Hence, the celebrated positive real lemma for standard state space systems is established for descriptor systems. In addition, the lossless positive realness of both standard state space systems and descriptor systems are characterized explicitly based on the controllable staircase ...

Linear descriptor systems are governed by dynamical equations subject to algebraic constraints. In the one-dimensional case, where only depend on a single index, usually time, Weierstrass canonical form splits up state vector in two parts, causal part, running forward and non-causal backward. this paper linear time-invariant autonomous two-dimensions discussed condition existence of non-trivial...

This paper presents a novel method for interest region description. We adopted the idea that the appearance of an interest region can be well characterized by the distribution of its local features. The most well-known descriptor built on this idea is the SIFT descriptor that uses gradient as the local feature. Thus far, existing texture features are not widely utilized in the context of region...

In this paper a contour matching based face recognition system is proposed, which uses “Fourier Descriptor of contour” for identification of faces. The feasibility of using contour and FD’s matching for human face identification is presented through experiment. The Fourier descriptor is used to describe the boundary of a shape in 2 dimensional space.The advantage of using the Fourier descriptor...

automatic facial recognition has many potential applications in different areas of humancomputer interaction. however, they are not yet fully realized due to the lack of an effectivefacial feature descriptor. in this paper, we present a new appearance based feature descriptor,the local directional pattern (ldp), to represent facial geometry and analyze its performance inrecognition. an ldp feat...

Since more than 50 years texture in image material is a topic of research. Hereby, color was ignored mostly. This study compares 70 different configurations for texture analysis, using four features. For the configurations we used: (i) a gray value texture descriptor: the co-occurrence matrix and a color texture descriptor: the color correlogram, (ii) six color spaces, and (iii) several quantiz...

A new molecular descriptor, nConf20, based on chemical connectivity, is presented which captures the accessible conformational space of a molecule. Currently the best available two-dimensional descriptors for quantifying the flexibility of a particular molecule are the rotatable bond count (RBC) and the Kier flexibility index. We present a descriptor which captures this information by sampling ...