We propose a novel MRF-based model for deformable image matching (also known as registration). The deformation is described by a field of discrete variables, representing displace...
In this paper, we develop a model for representing term dependence based on Markov Random Fields and present an approach based on Markov Chain Monte Carlo technique for generating ...
We address the normal reconstruction problem by photometric stereo using a uniform and dense set of photometric images captured at fixed viewpoint. Our method is robust to spurio...
Abstract. Peptide mass fingerprinting is a technique to identify a protein from its fragment masses obtained by mass spectrometry after enzymatic digestion. Recently, much attentio...
Abstract –In this paper, a variational message passing framework is proposed for Markov random fields, which is computationally more efficient and admits wider applicability comp...