Automatically Selecting Inference Algorithms for Discrete Energy Minimisation

Paul Henderson; Vittorio Ferrari

doi:10.1007/978-3-319-46454-1_15

Automatically Selecting Inference Algorithms for Discrete Energy Minimisation

Paul Henderson, Vittorio Ferrari

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract / Description of output

Minimisation of discrete energies defined over factors is an important problem in computer vision, and a vast number of MAP inference algorithms have been proposed. Different inference algorithms perform better on factor graph models (GMs) from different underlying problem classes, and in general it is difficult to know which algorithm will yield the lowest energy for a given GM. To mitigate this difficulty, survey papers [1–3] advise the practitioner on what algorithms perform well on what classes of models. We take the next step forward, and present a technique to automatically select the best inference algorithm for an input GM. We validate our method experimentally on an extended version of the OpenGM2 benchmark [3], containing a diverse set of vision problems. On average, our method selects an inference algorithm yielding labellings with 96 % of variables the same as the best available algorithm.

Original language	English
Title of host publication	Computer Vision -- ECCV 2016
Subtitle of host publication	14th European Conference, Amsterdam, The Netherlands, October 11-14, 2016, Proceedings, Part V
Editors	Bastian Leibe, Jiri Matas, Nicu Sebe, Max Welling
Place of Publication	Cham
Publisher	Springer
Pages	235-252
Number of pages	18
ISBN (Electronic)	978-3-319-46454-1
ISBN (Print)	978-3-319-46453-4
DOIs	https://doi.org/10.1007/978-3-319-46454-1_15
Publication status	Published - 16 Sept 2016
Event	14th European Conference on Computer Vision 2016 - Amsterdam, Netherlands Duration: 8 Oct 2016 → 16 Oct 2016 http://www.eccv2016.org/

Publication series

Name	Lecture Notes in Computer Science
Publisher	Springer International Publishing
Volume	9909
ISSN (Print)	0302-9743

Conference

Conference	14th European Conference on Computer Vision 2016
Abbreviated title	ECCV 2016
Country/Territory	Netherlands
City	Amsterdam
Period	8/10/16 → 16/10/16
Internet address	http://www.eccv2016.org/

Access to Document

10.1007/978-3-319-46454-1_15

0970Accepted author manuscript, 2.43 MB

http://link.springer.com/chapter/10.1007%2F978-3-319-46454-1_15

Cite this

Henderson, P., & Ferrari, V. (2016). Automatically Selecting Inference Algorithms for Discrete Energy Minimisation. In B. Leibe, J. Matas, N. Sebe, & M. Welling (Eds.), Computer Vision -- ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11-14, 2016, Proceedings, Part V (pp. 235-252). (Lecture Notes in Computer Science; Vol. 9909). Springer. https://doi.org/10.1007/978-3-319-46454-1_15

Henderson, Paul ; Ferrari, Vittorio. / Automatically Selecting Inference Algorithms for Discrete Energy Minimisation. Computer Vision -- ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11-14, 2016, Proceedings, Part V. editor / Bastian Leibe ; Jiri Matas ; Nicu Sebe ; Max Welling. Cham : Springer, 2016. pp. 235-252 (Lecture Notes in Computer Science).

@inproceedings{7cce4d761fe34a46ab73b00ad0787cae,

title = "Automatically Selecting Inference Algorithms for Discrete Energy Minimisation",

abstract = "Minimisation of discrete energies defined over factors is an important problem in computer vision, and a vast number of MAP inference algorithms have been proposed. Different inference algorithms perform better on factor graph models (GMs) from different underlying problem classes, and in general it is difficult to know which algorithm will yield the lowest energy for a given GM. To mitigate this difficulty, survey papers [1–3] advise the practitioner on what algorithms perform well on what classes of models. We take the next step forward, and present a technique to automatically select the best inference algorithm for an input GM. We validate our method experimentally on an extended version of the OpenGM2 benchmark [3], containing a diverse set of vision problems. On average, our method selects an inference algorithm yielding labellings with 96 % of variables the same as the best available algorithm.",

author = "Paul Henderson and Vittorio Ferrari",

year = "2016",

month = sep,

day = "16",

doi = "10.1007/978-3-319-46454-1_15",

language = "English",

isbn = "978-3-319-46453-4",

series = "Lecture Notes in Computer Science",

publisher = "Springer",

pages = "235--252",

editor = "Bastian Leibe and Jiri Matas and Nicu Sebe and Max Welling",

booktitle = "Computer Vision -- ECCV 2016",

address = "United Kingdom",

note = "14th European Conference on Computer Vision 2016, ECCV 2016 ; Conference date: 08-10-2016 Through 16-10-2016",

url = "http://www.eccv2016.org/",

}

Henderson, P & Ferrari, V 2016, Automatically Selecting Inference Algorithms for Discrete Energy Minimisation. in B Leibe, J Matas, N Sebe & M Welling (eds), Computer Vision -- ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11-14, 2016, Proceedings, Part V. Lecture Notes in Computer Science, vol. 9909, Springer, Cham, pp. 235-252, 14th European Conference on Computer Vision 2016, Amsterdam, Netherlands, 8/10/16. https://doi.org/10.1007/978-3-319-46454-1_15

Automatically Selecting Inference Algorithms for Discrete Energy Minimisation. / Henderson, Paul; Ferrari, Vittorio.
Computer Vision -- ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11-14, 2016, Proceedings, Part V. ed. / Bastian Leibe; Jiri Matas; Nicu Sebe; Max Welling. Cham: Springer, 2016. p. 235-252 (Lecture Notes in Computer Science; Vol. 9909).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Automatically Selecting Inference Algorithms for Discrete Energy Minimisation

AU - Henderson, Paul

AU - Ferrari, Vittorio

PY - 2016/9/16

Y1 - 2016/9/16

N2 - Minimisation of discrete energies defined over factors is an important problem in computer vision, and a vast number of MAP inference algorithms have been proposed. Different inference algorithms perform better on factor graph models (GMs) from different underlying problem classes, and in general it is difficult to know which algorithm will yield the lowest energy for a given GM. To mitigate this difficulty, survey papers [1–3] advise the practitioner on what algorithms perform well on what classes of models. We take the next step forward, and present a technique to automatically select the best inference algorithm for an input GM. We validate our method experimentally on an extended version of the OpenGM2 benchmark [3], containing a diverse set of vision problems. On average, our method selects an inference algorithm yielding labellings with 96 % of variables the same as the best available algorithm.

AB - Minimisation of discrete energies defined over factors is an important problem in computer vision, and a vast number of MAP inference algorithms have been proposed. Different inference algorithms perform better on factor graph models (GMs) from different underlying problem classes, and in general it is difficult to know which algorithm will yield the lowest energy for a given GM. To mitigate this difficulty, survey papers [1–3] advise the practitioner on what algorithms perform well on what classes of models. We take the next step forward, and present a technique to automatically select the best inference algorithm for an input GM. We validate our method experimentally on an extended version of the OpenGM2 benchmark [3], containing a diverse set of vision problems. On average, our method selects an inference algorithm yielding labellings with 96 % of variables the same as the best available algorithm.

U2 - 10.1007/978-3-319-46454-1_15

DO - 10.1007/978-3-319-46454-1_15

M3 - Conference contribution

SN - 978-3-319-46453-4

T3 - Lecture Notes in Computer Science

SP - 235

EP - 252

BT - Computer Vision -- ECCV 2016

A2 - Leibe, Bastian

A2 - Matas, Jiri

A2 - Sebe, Nicu

A2 - Welling, Max

PB - Springer

CY - Cham

T2 - 14th European Conference on Computer Vision 2016

Y2 - 8 October 2016 through 16 October 2016

ER -

Henderson P, Ferrari V. Automatically Selecting Inference Algorithms for Discrete Energy Minimisation. In Leibe B, Matas J, Sebe N, Welling M, editors, Computer Vision -- ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11-14, 2016, Proceedings, Part V. Cham: Springer. 2016. p. 235-252. (Lecture Notes in Computer Science). doi: 10.1007/978-3-319-46454-1_15

Automatically Selecting Inference Algorithms for Discrete Energy Minimisation

Abstract / Description of output

Publication series

Conference

Access to Document

Fingerprint

Cite this