Fusion Moves for Graph Matching

Hutschenreiter, Lisa; Haller, Stefan; Feineis, Lorenz; Rother, Carsten; Kainmüller, Dagmar; Savchynskyy, Bogdan

doi:10.1109/iccv48922.2021.00621

Cited by 9 publications

(36 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our work significantly excels evaluations in all the papers introducing the algorithms we study. This implies also to the largest existing comparison [31] so far. The latter considers only 8 out of the 11 datasets and evaluates 6 algorithms out of our 20.…”

Section: Contributionsmentioning

confidence: 77%

“…The graph matching problem can be represented in the form of a maximum a posteriori (MAP) inference problem for discrete graphical models [54], known also as Markov random field (MRF) energy minimization and closely related to valued and weighted constraint satisfaction problems. As several graph matching works in computer vision [31,57,67], use this representation, we provide it below in more detail.…”

Section: Path Followingmentioning

confidence: 99%

“…Randomized generation and fusion based. Fusion moves with a greedy heuristic (fm) [31] is based on the graphical model representation and consists of two parts: A randomized greedy assignment generation, and fusion of the assignments. The randomized generator greedily fixes labels in the nodes in a way that minimizes the objective value restricted to the already fixed labels.…”

Section: Path Following Basedmentioning

confidence: 99%

“…The methods below consider the Lagrange decompositions [28] of the graph matching problem (1) [59], or its graphical model representation (3) [31,57,67], and optimize the corresponding dual. The methods differ in the dual optimization and chosen primal solution reconstruction algorithms.…”

Section: Lagrange Duality-based Techniquesmentioning

confidence: 99%

See 3 more Smart Citations

A Comparative Study of Graph Matching Algorithms in Computer Vision

Haller¹,

Feineis²,

Hutschenreiter³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

The graph matching optimization problem is an essential component for many tasks in computer vision, such as bringing two deformable objects in correspondence. Naturally, a wide range of applicable algorithms have been proposed in the last decades. Since a common standard benchmark has not been developed, their performance claims are often hard to verify as evaluation on differing problem instances and criteria make the results incomparable. To address these shortcomings, we present a comparative study of graph matching algorithms. We create a uniform benchmark where we collect and categorize a large set of existing and publicly available computer vision graph matching problems in a common format. At the same time we collect and categorize the most popular open-source implementations of graph matching algorithms. Their performance is evaluated in a way that is in line with the best practices for comparing optimization algorithms. The study is designed to be reproducible and extensible to serve as a valuable resource in the future.Our study provides three notable insights: (i) popular problem instances are exactly solvable in substantially less than 1 second, and, therefore, are insufficient for future empirical evaluations; (ii) the most popular baseline methods are highly inferior to the best available methods; (iii) despite the NP-hardness of the problem, instances coming from vision applications are often solvable in a few seconds even for graphs with more than 500 vertices.

show abstract

Section: Contributionsmentioning

confidence: 77%

Section: Path Followingmentioning

confidence: 99%

Section: Path Following Basedmentioning

confidence: 99%

Section: Lagrange Duality-based Techniquesmentioning

confidence: 99%

See 2 more Smart Citations

A Comparative Study of Graph Matching Algorithms in Computer Vision

Haller¹,

Feineis²,

Hutschenreiter³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…While we are aware of even faster purely primal heuristics [10,35] for MRF and e.g. [26] for graph matching they do not optimize a convex relaxation and hence do not provide lower bounds. Hence, we have chosen TRWS [33] for MRF and AMP [51] for graph matching which, similar to FastDOG, optimize an equivalent resp.…”

Section: Gurobimentioning

confidence: 99%

FastDOG: Fast Discrete Optimization on GPU

Abbas¹,

Swoboda²

2021

Preprint

View full text Add to dashboard Cite

We present a massively parallel Lagrange decomposition method for solving 0-1 integer linear programs occurring in structured prediction. We propose a new iterative update scheme for solving the Lagrangean dual and a perturbation technique for decoding primal solutions. For representing subproblems we follow [37] and use binary decision diagrams (BDDs). Our primal and dual algorithms require little synchronization between subproblems and optimization over BDDs needs only elementary operations without complicated control flow. This allows us to exploit the parallelism offered by GPUs for all components of our method. We present experimental results on combinatorial problems from MAP inference for Markov Random Fields, quadratic assignment and cell tracking for developmental biology. Our highly parallel GPU implementation improves upon the running times of the algorithms from [37] by up to an order of magnitude. In particular, we come close to or outperform some state-of-the-art specialized heuristics while being problem agnostic.

show abstract