Stochastic image modeling based on conventional Markov random fields is extensively discussed in the literature. A new stochastic image model based on Markov random fields is intr...
In contrast to traditional Markov random field (MRF) models, we develop a Steerable Random Field (SRF) in which the field potentials are defined in terms of filter responses that ...
Contextual reasoning through graphical models such as Markov Random Fields often show superior performance against local classifiers in many domains. Unfortunately, this performanc...
Drawbacks of the traditional scenario of image modeling by Gibbs random fields with multiple pairwise pixel interactions are outlined, and a more reasonable alternative scenario b...