Re-writing most robust estimation schemes through the lens of covariance estimation when attempting the find the set of states that maximize the a posteriori estimate.