This paper presents a generative model based approach to man-made structure detection in 2D natural images. The proposed approach uses a causal multiscale random field suggested i...
We seek to both detect and segment objects in images. To exploit both local image data as well as contextual information, we introduce Boosted Random Fields (BRFs), which use boos...
Antonio Torralba, Kevin P. Murphy, William T. Free...
3D Bayesian regularization applied to diffusion tensor MRI is presented here. The approach uses Markov Random Field ideas and is based upon the definition of a 3D neighborhood syst...
Abstract. In this paper, we propose an unsupervised color image segmentation scheme using homotopy continuation method and Compound Markov Random Field (CMRF) model. The proposed s...
The use of random fields, which allows one to take into account the spatial interaction among random variables in complex systems, is a frequent tool in numerous problems of stati...