Gauss Mixture Model Clustering for Noisy Images under Rate Constraints

Abstract

We consider the problem of classification based on Gauss mixture models for a simple network of two sensors with noisy observations. The goal of each sensor is to give a classification decision based on its noisy observation and the help it receives from the other sensor under the given rate constraint. We formulate the problem as a vector quantization problem and design a Lloyd optimal quantizer, minimizing the classification error for the given rate constraint. Our cross-validated simulations, using a set of aerial images, indicate an improvement in the classification performance (for the given rate constraints) when compared with simple extensions of previously published GMM-based algorithms.