Maximum mean discrepancy (MMD) has been successfully applied to learn deep generative models for characterizing a joint distribution of variables via kernel mean embedding. In this paper, we present conditional generative moment-matching networks (CGMMN), which learn a conditional distribution given some input variables based on a conditional maximum mean discrepancy (CMMD) criterion. The learn...

This paper proposes a method for matching images based on their higher order moments without knowing the point correspondences. It is assumed that the disparity between the images can be explained by an affine transformation. The second order statistics is used to transform the image points into canonical form, which reduces the affine matching problem for determining an orthonormal transformat...

The problem of proposing a framework for linear moment matching is tackled. It is possible to define five notions of moments, i.e., two from a complex ”s-domain” point of view and three from a time-domain point of view. Based on these notions, reduced order models that achieve moment matching are obtained. This paper sets the basis to develop the theory of model reduction by moment matching for...

Problem statement: Template matching is a famous methodology that has a wide range of applications in image and signal processing. For a template and input image, template matching methodology finds the partial input image that is most closely matches the template image in terms of specific criterion such as the Euclidean distance or cross-correlation. Approach: In this study, a fast and robust...

In this paper we propose a solution to the problem of moment matching with preservation of the port Hamiltonian structure, in the framework of time-domain moment matching. We characterize several families of parameterized port Hamiltonian models that match the moments of a given port Hamiltonian system, at a set of finite interpolation points. We also discuss the problem of Markov parameters ma...