Paper Info
Title
Efficient Continuous Relaxations For Dense Crf
Abstract
Dense conditional random fields (CRF) with Gaussian pairwise potentials have emerged as a popular framework for several computer vision applications such as stereo correspondence and semantic segmentation. By modeling long-range interactions, dense CRFs provide a more detailed labelling compared to their sparse counterparts. Variational inference in these dense models is performed using a filtering-based mean-field algorithm in order to obtain a fully-factorized distribution minimising the Kullback-Leibler divergence to the true distribution. In contrast to the continuous relaxation-based energy minimisation algorithms used for sparse CRFs, the mean-field algorithm fails to provide strong theoretical guarantees on the quality of its solutions. To address this deficiency, we show that it is possible to use the same filtering approach to speed-up the optimisation of several continuous relaxations. Specifically, we solve a convex quadratic programming (QP) relaxation using the efficient Frank-Wolfe algorithm. This also allows us to solve difference-of-convex relaxations via the iterative concave-convex procedure where each iteration requires solving a convex QP. Finally, we develop a novel divide-and-conquer method to compute the subgradients of a linear programming relaxation that provides the best theoretical bounds for energy minimisation. We demonstrate the advantage of continuous relaxations over the widely used mean-field algorithm on publicly available datasets.
Year
DOI
Venue
2016
10.1007/978-3-319-46475-6_50
COMPUTER VISION - ECCV 2016, PT II
Keywords
DocType
Volume
Energy minimisation, Dense CRF, Inference, Linear programming, Quadratic programming
Conference
9906
ISSN
Citations 
PageRank 
0302-9743
3
0.43
References 
Authors
20
5
Name
Order
Citations
PageRank
Alban Desmaison1174.63
Rudy Bunel2405.28
Pushmeet Kohli37398332.84
Philip H. S. Torr49140636.18
M. Pawan Kumar5102382.37