Inference and Learning of Graphical Models: Theory and Applications in Computer Vision and Image Analysis

摘要

Graphical models have become a ubiquitous methodology for computer vision and image analysis problems, in terms of both the expressive potential of the modeling process and the optimal-ity properties of the corresponding inference algorithms. A variety of graphical models as well as inference and learning methods have been developed for addressing numerous low, mid and high-level vision problems. The main stream referred to pairwise models. Whereas, we have also witnessed significant progress in higher-order models during recent several years, which substantially enhances the model expressiveness and expands the domain of solvable problems.