Structured Robust Covariance Estimation

Considers the estimation of covariance matrices in non-standard conditions including heavy-tailed distributions and outlier contamination. Prior knowledge on the structure of these matrices is exploited in order to improve the estimation accuracy.