Fast multi-labelling in early vision

Abstract

Multi-labelling algorithms are widely used in solving early vision problems. These early vision problems, which include image segmentation, stereo correspondence estimation, etc., are of fundamental importance in computer vision research and applications. Solving a multi-label problem is in general NP-hard. Despite the notable success in the development of multi-labelling algorithms in recent years, heavy computation is still required especially when the size of the problem becomes large, which makes multi-labelling in general a slow process. The limitation in efficiency bottlenecks the availability of multi-labelling algorithms for practical applications and scientific research. This thesis aims to improve the situation via significantly elevating the solving speed of multi-labelling problems in early vision applications. - provided by Candidate.