Principled Parallel Mean-Field Inference for Discrete Random Fields

Baqué, Pierre; Bagautdinov, Timur; Fleuret, François; Fua, Pascal

doi:10.1109/cvpr.2016.630

Cited by 19 publications

(24 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the parallel scheme, we update all the probabilities Q x (s x = l) in parallel for each x and l (Equation (10)). Although naive parallelization does not guarantee convergence to a local minimum, Baqué et al [BBFF16] showed how using additional damping in each fix point step would guarantee convergence. In our case, however, we did not experience the need for such a damping factor.…”

Section: Parallel Mean-field Inference On the Gpumentioning

confidence: 99%

Temporally Consistent Motion Segmentation From RGB‐D Video

Bertholet

Ichim

Zwicker

2018

Computer Graphics Forum

View full text Add to dashboard Cite

Temporally consistent motion segmentation from RGB‐D videos is challenging because of the limitations of current RGB‐D sensors. We formulate segmentation as a motion assignment problem, where a motion is a sequence of rigid transformations through all frames of the input. We capture the quality of each potential assignment by defining an appropriate energy function that accounts for occlusions and a sensor‐specific noise model. To make energy minimization tractable, we work with a discrete set instead of the continuous, high dimensional space of motions, where the discrete motion set provides an upper bound for the original energy. We repeatedly minimize our energy, and in each step extend and refine the motion set to further lower the bound. A quantitative comparison to the current state of the art demonstrates the benefits of our approach in difficult scenarios.

show abstract

Section: Parallel Mean-field Inference On the Gpumentioning

confidence: 99%

Temporally Consistent Motion Segmentation From RGB‐D Video

Bertholet

Ichim

Zwicker

2018

Computer Graphics Forum

View full text Add to dashboard Cite

show abstract

“…Finally, it is worth mentioning that total-variation (TV) terms can be viewed as the continuous counterpart of Potts regulariza-tion, and were the subject of a large number of vision works in recent years [7,23,24,30]. Recently, there have been significant research efforts focusing on designing parallel (or distributed) formulations for optimizing pairwise functions [1,19,25]. Distributing computations would be beneficial not only to highresolution images and massive 3D grids but also to difficult high-order models [11,14,26], which require approximate solutions solving a large number of problems of the form (1).…”

Section: Introductionmentioning

confidence: 99%

“…Unfortunately such techniques are restricted to TV regularization terms. Mean-field inference techniques [1,16] also attracted significant attention recently as they can be parrallelized, albeit at the cost of convergence guarantees [1].…”

Section: Introductionmentioning

confidence: 99%

DOPE: Distributed Optimization for Pairwise Energies

Dolz

Ayed

Desrosiers

2017

2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)

View full text Add to dashboard Cite

We formulate an Alternating Direction Method of Multipliers (ADMM) that systematically distributes the computations of any technique for optimizing pairwise functions, including non-submodular potentials. Such discrete functions are very useful in segmentation and a breadth of other vision problems. Our method decomposes the problem into a large set of small sub-problems, each involving a subregion of the image domain, which can be solved in parallel. We achieve consistency between the sub-problems through a novel constraint that can be used for a large class of pairwise functions. We give an iterative numerical solution that alternates between solving the sub-problems and updating consistency variables, until convergence. We report comprehensive experiments, which demonstrate the benefit of our general distributed solution in the case of the popular serial algorithm of Boykov and Kolmogorov (BK algorithm) and, also, in the context of non-submodular functions.

show abstract

“…While the initial work by [19] used a version of mean-field that is not guaranteed to converge, their follow-up paper [20] proposed a convergent mean-field algorithm for negative semi-definite label compatibility functions. Recently, Baqué et al [4] presented a new algorithm that has convergence guarantees in the general case. Vineet et al [34] extended the mean-field model to allow the addition of higher-order terms on top of the dense pairwise potentials, enabling the use of co-occurence potentials [24] and P n -Potts models [15].…”

mentioning

confidence: 99%

“…Our work can be viewed as a complementary direction to previous research trends in dense CRFs. While [4,20,34] improved mean-field and [29,38] learnt the parameters, we focus on the energy minimisation problem.…”

mentioning

confidence: 99%

Efficient Relaxations for Dense CRFs with Sparse Higher-Order Potentials

Joy¹,

Desmaison²,

Ajanthan³

et al. 2019

SIAM J. Imaging Sci.

View full text Add to dashboard Cite

Dense conditional random fields (CRFs) have become a popular framework for modelling several problems in computer vision such as stereo correspondence and multi-class semantic segmentation. By modelling longrange interactions, dense CRFs provide a labelling that captures finer detail than their sparse counterparts. Currently, the state-of-the-art algorithm performs mean-field inference using a filter-based method but fails to provide a strong theoretical guarantee on the quality of the solution.A question naturally arises as to whether it is possible to obtain a maximum a posteriori (MAP) estimate of a dense CRF using a principled method. Within this paper, we show that this is indeed possible. We will show that, by using a filter-based method, continuous relaxations of the MAP problem can be optimised efficiently using state-of-the-art algorithms. Specifically, we will solve a quadratic programming (QP) relaxation using the Frank-Wolfe algorithm and a linear programming (LP) relaxation by developing a proximal minimisation framework. By exploiting labelling consistency in the higher-order potentials and utilising the filter-based method, we are able to formulate the above algorithms such that each iteration has a complexity linear in the number of classes and random variables. The presented algorithms can be applied to any labelling problem using a dense CRF with sparse higher-order potentials. In this paper, we use semantic segmentation as an example application as it demonstrates the ability of the algorithm to scale to dense CRFs with large dimensions. We perform experiments on the Pascal dataset to indicate that the presented algorithms are able to attain lower energies than the mean-field inference method. * Indicates equal contribution arXiv:1805.09028v2 [cs.CV]

show abstract

Principled Parallel Mean-Field Inference for Discrete Random Fields

Cited by 19 publications

References 31 publications

Temporally Consistent Motion Segmentation From RGB‐D Video

Temporally Consistent Motion Segmentation From RGB‐D Video

DOPE: Distributed Optimization for Pairwise Energies

Efficient Relaxations for Dense CRFs with Sparse Higher-Order Potentials

Contact Info

Product

Resources

About