Approximate Parameter Learning in Conditional Random Fields: An Empirical Investigation

摘要

We investigate maximum likelihood parameter learning in Conditional Random Fields (CRF) and present an empirical study of pseudo-likelihood (PL) based approximations of the parameter likelihood gradient. We show, as opposed to [1][2], that these parameter learning methods can be improved and evaluate the resulting performance employing different inference techniques. We show that the approximation based on penalized pseudo-likelihood (PPL) in combination with the Maximum A Posteriori (MAP) inference yields results comparable to other state of the art approaches, while providing advantages to formulating parameter learning as a convex optimization problem. Eventually, we demonstrate applicability on the task of detecting man-made structures in natural images.