Unless there are specific questions about the Gibbs estimator I'm just going to skip the estimation part and move along to think about how to parameterize this prior, then I want to get into the empirical results. The details can be found in Appendix A of the paper.
Go to written version of paper