What is the prerequisite knowledge to *really* "get" conditional random fields, HMMs, etc?

I do work in computer vision, but I have avoided these topics for a while since the Bayesian probability theory doesn't always seem to make sense to me in the context of computer vision.

I want to understand it right down to the theory. I'm going to have some off time over the holidays that I will do some reading in. If I want this stuff to be entirely concrete to me, what should I be reading? It doesn't seem like starting right from the initial CRF paper is the right approach :)