Abstract | ||
---|---|---|
The current paper introduces new prior distributions on the zero-mean multivariate Gaussian model, with the aim of applying them to the classification of covariance matrices populations. These new prior distributions are entirely based on the Riemannian geometry of the multivariate Gaussian model. More precisely, the proposed Riemannian Gaussian distribution has two parameters, the centre of mass (Y) over bar and the dispersion parameter sigma. Its density with respect to Riemannian volume is proportional to exp(-d(2) (Y; (Y) over bar)), where d(2) (Y; (Y) over bar) is the square of Rao's Riemannian distance. We derive its maximum likelihood estimators and propose an experiment on the VisTex database for the classification of texture images. |
Year | DOI | Venue |
---|---|---|
2015 | 10.1007/978-3-319-25040-3_40 | GEOMETRIC SCIENCE OF INFORMATION, GSI 2015 |
Keywords | DocType | Volume |
Texture classification, Information geometry, Riemannian centre of mass, Mixture estimation, EM algorithm | Conference | 9389 |
ISSN | Citations | PageRank |
0302-9743 | 5 | 0.44 |
References | Authors | |
2 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Salem Said | 1 | 11 | 2.23 |
Lionel Bombrun | 2 | 150 | 20.59 |
Y. Berthoumieu | 3 | 389 | 51.66 |