Energy minimization methods in computer vision and pattern recognition

Más información sobre el libro

The content covers a range of advanced topics in discrete optimization and Markov random fields. It explores multi-label moves for MRFs with truncated convex priors, detection and segmentation of independently moving objects from dense scene flow, and efficient global minimization for the multiphase Chan-Vese model of image segmentation. Techniques such as bipartite graph matching on GPU and pose-invariant face matching using MRF energy minimization are discussed. The text also delves into parallel hidden hierarchical fields for multi-scale reconstruction and general search algorithms for energy minimization problems, alongside partial differential equations. Additional topics include complex diffusion on scalar and vector-valued image graphs, coupled super-resolution with non-parametric motion, and optical flow decomposition models. Various approaches to image segmentation and tracking are presented, including hierarchical pairwise segmentation and tracking as segmentation of spatial-temporal volumes. The annotation also covers parameter estimation for marked point processes in object extraction from remote sensing images and robust segmentation techniques. Furthermore, it addresses shape optimization and registration, intrinsic second-order geometric optimization, and image registration under varying illumination. Techniques for inpainting and image denoising, exemplar-based interpolation, and color image restoration using