Ying Nian Wu University of California, Los Angeles (UCLA) Statistics This lecture explains the basics of the Markov random field (MRF) models, including their origin in statistical physics and early work on MRF in statistics. I will then describe the Markov chain Monte Carlo (MCMC) algorithms for sampling from such models, including the Gibbs sampler, the Metropolis algorithm and the hybrid (Hamiltonian) Monte Carlo algorithm etc. After that I will discuss the algorithms for learning the parameters of the MRF models. Finally I will describe the applications of MRF models in computer vision, including image segmentation and texture modeling.