Self-Organizing Mixture Networks for Representation of Grayscale Digital Images

TL;DR

This paper introduces Self-Organizing Mixture Networks (SOMN) as an unsupervised method to represent grayscale images by approximating them with Gaussian mixtures, enhancing image modeling techniques.

Contribution

It proposes a novel application of SOMN for representing grayscale images as Gaussian mixtures, extending the use of self-organizing maps in image processing.

Findings

SOMN effectively approximates grayscale images as Gaussian mixtures.

The method provides an unsupervised approach for image representation.

Potential for improved image analysis and compression.

Abstract

Self-Organizing Maps are commonly used for unsupervised learning purposes. This paper is dedicated to the certain modification of SOM called SOMN (Self-Organizing Mixture Networks) used as a mechanism for representing grayscale digital images. Any grayscale digital image regarded as a distribution function can be approximated by the corresponding Gaussian mixture. In this paper, the use of SOMN is proposed in order to obtain such approximations for input grayscale images in unsupervised manner.