1: \begin{abstract}

2: %\baselineskip0.25in

3: Due to its heavy-tailed and fully parametric form, the multivariate generalized Gaussian distribution (MGGD) has been receiving much attention for modeling extreme events in signal and image processing applications. Considering the estimation issue of the MGGD parameters, the main contribution of this paper is to prove that the maximum likelihood estimator (MLE) of the scatter matrix exists and is unique up to a scalar factor, for a given shape parameter $\beta\in(0,1)$. Moreover, an estimation algorithm based on a Newton-Raphson recursion is proposed for computing the MLE of MGGD parameters. Various experiments conducted on synthetic and real data are presented to illustrate the theoretical derivations in terms of number of iterations and number of samples for different values of the shape parameter. The main conclusion of this work is that the parameters of MGGDs can be estimated using the maximum likelihood principle with good performance.

4: 

5: %Due to its more peaky and heavy-tailed shape compared to the multivariate Gaussian distribution, the multivariate generalized Gaussian distribution has recently been proposed as a relevant model for several image processing applications. Unfortunately, the statistical properties of the maximum likelihood estimators have not yet been investigated. The main paper contribution is to prove that for the shape parameter $\beta\in]0,1[$ (which corresponds to most of the real-life problems), the maximum likelihood estimator of the scatter matrix exists and is unique up to a scalar factor. An estimation algorithm based on a Newton-Raphson recursion is investigated. Some experiments illustrate the convergence speed of this algorithm. The bias and consistency of the scatter matrix MLE are then studied for different values of the shape parameter. The performance of the shape parameter estimator is finally addressed by comparing its variance to the Cram\'er-Rao bound.

6: \end{abstract}

7: