Weakly Supervised Learning of Foreground-Background Segmentation using Masked RBMs

Nicolas Heess; Nicolas Le Roux; John Winn

doi:10.1007/978-3-642-21738-8_2

Weakly Supervised Learning of Foreground-Background Segmentation using Masked RBMs

Nicolas Heess, Nicolas Le Roux, John Winn

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

Abstract

We propose an extension of the Restricted Boltzmann Machine (RBM) that allows the joint shape and appearance of foreground objects in cluttered images to be modeled independently of the background. We present a learning scheme that learns this representation directly from cluttered images with only very weak supervision. The model generates plausible samples and performs foreground-background segmentation. We demonstrate that representing foreground objects independently of the background can be beneficial in recognition tasks.

Original language	English
Title of host publication	Artificial Neural Networks and Machine Learning – ICANN 2011
Subtitle of host publication	21st International Conference on Artificial Neural Networks, Espoo, Finland, June 14-17, 2011, Proceedings, Part II
Publisher	Springer
Pages	9-16
Number of pages	8
ISBN (Electronic)	978-3-642-21738-8
ISBN (Print)	978-3-642-21737-1
DOIs	https://doi.org/10.1007/978-3-642-21738-8_2
Publication status	Published - 2011
Event	International Conference on Artificial Neural Networks (ICANN) - Espoo, Finland Duration: 14 Jun 2011 → 17 Jun 2011

Publication series

Name	Lecture Notes in Computer Science
Publisher	Springer Berlin Heidelberg
Volume	6792
ISSN (Print)	0302-9743

Conference

Conference	International Conference on Artificial Neural Networks (ICANN)
Country/Territory	Finland
City	Espoo
Period	14/06/11 → 17/06/11

Access to Document

10.1007/978-3-642-21738-8_2

http://link.springer.com/chapter/10.1007%2F978-3-642-21738-8_2

Cite this

Heess, N., Le Roux, N., & Winn, J. (2011). Weakly Supervised Learning of Foreground-Background Segmentation using Masked RBMs. In Artificial Neural Networks and Machine Learning – ICANN 2011: 21st International Conference on Artificial Neural Networks, Espoo, Finland, June 14-17, 2011, Proceedings, Part II (pp. 9-16). (Lecture Notes in Computer Science; Vol. 6792). Springer. https://doi.org/10.1007/978-3-642-21738-8_2

Heess, Nicolas ; Le Roux, Nicolas ; Winn, John. / Weakly Supervised Learning of Foreground-Background Segmentation using Masked RBMs. Artificial Neural Networks and Machine Learning – ICANN 2011: 21st International Conference on Artificial Neural Networks, Espoo, Finland, June 14-17, 2011, Proceedings, Part II. Springer, 2011. pp. 9-16 (Lecture Notes in Computer Science).

@inproceedings{c1b9e3b966b94f1f8f8dc028c841b557,

title = "Weakly Supervised Learning of Foreground-Background Segmentation using Masked RBMs",

abstract = "We propose an extension of the Restricted Boltzmann Machine (RBM) that allows the joint shape and appearance of foreground objects in cluttered images to be modeled independently of the background. We present a learning scheme that learns this representation directly from cluttered images with only very weak supervision. The model generates plausible samples and performs foreground-background segmentation. We demonstrate that representing foreground objects independently of the background can be beneficial in recognition tasks.",

author = "Nicolas Heess and \{Le Roux\}, Nicolas and John Winn",

year = "2011",

doi = "10.1007/978-3-642-21738-8\_2",

language = "English",

isbn = "978-3-642-21737-1",

series = "Lecture Notes in Computer Science",

publisher = "Springer",

pages = "9--16",

booktitle = "Artificial Neural Networks and Machine Learning – ICANN 2011",

address = "United Kingdom",

note = "International Conference on Artificial Neural Networks (ICANN) ; Conference date: 14-06-2011 Through 17-06-2011",

}

Heess, N, Le Roux, N & Winn, J 2011, Weakly Supervised Learning of Foreground-Background Segmentation using Masked RBMs. in Artificial Neural Networks and Machine Learning – ICANN 2011: 21st International Conference on Artificial Neural Networks, Espoo, Finland, June 14-17, 2011, Proceedings, Part II. Lecture Notes in Computer Science, vol. 6792, Springer, pp. 9-16, International Conference on Artificial Neural Networks (ICANN), Espoo, Finland, 14/06/11. https://doi.org/10.1007/978-3-642-21738-8_2

Weakly Supervised Learning of Foreground-Background Segmentation using Masked RBMs. / Heess, Nicolas; Le Roux, Nicolas; Winn, John.
Artificial Neural Networks and Machine Learning – ICANN 2011: 21st International Conference on Artificial Neural Networks, Espoo, Finland, June 14-17, 2011, Proceedings, Part II. Springer, 2011. p. 9-16 (Lecture Notes in Computer Science; Vol. 6792).

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

TY - GEN

T1 - Weakly Supervised Learning of Foreground-Background Segmentation using Masked RBMs

AU - Heess, Nicolas

AU - Le Roux, Nicolas

AU - Winn, John

PY - 2011

Y1 - 2011

N2 - We propose an extension of the Restricted Boltzmann Machine (RBM) that allows the joint shape and appearance of foreground objects in cluttered images to be modeled independently of the background. We present a learning scheme that learns this representation directly from cluttered images with only very weak supervision. The model generates plausible samples and performs foreground-background segmentation. We demonstrate that representing foreground objects independently of the background can be beneficial in recognition tasks.

AB - We propose an extension of the Restricted Boltzmann Machine (RBM) that allows the joint shape and appearance of foreground objects in cluttered images to be modeled independently of the background. We present a learning scheme that learns this representation directly from cluttered images with only very weak supervision. The model generates plausible samples and performs foreground-background segmentation. We demonstrate that representing foreground objects independently of the background can be beneficial in recognition tasks.

U2 - 10.1007/978-3-642-21738-8_2

DO - 10.1007/978-3-642-21738-8_2

M3 - Conference contribution

SN - 978-3-642-21737-1

T3 - Lecture Notes in Computer Science

SP - 9

EP - 16

BT - Artificial Neural Networks and Machine Learning – ICANN 2011

PB - Springer

T2 - International Conference on Artificial Neural Networks (ICANN)

Y2 - 14 June 2011 through 17 June 2011

ER -

Heess N, Le Roux N, Winn J. Weakly Supervised Learning of Foreground-Background Segmentation using Masked RBMs. In Artificial Neural Networks and Machine Learning – ICANN 2011: 21st International Conference on Artificial Neural Networks, Espoo, Finland, June 14-17, 2011, Proceedings, Part II. Springer. 2011. p. 9-16. (Lecture Notes in Computer Science). doi: 10.1007/978-3-642-21738-8_2

Weakly Supervised Learning of Foreground-Background Segmentation using Masked RBMs

Abstract

Publication series

Conference

Access to Document

Fingerprint

Cite this