Deconvolutional networks

Matthew D. Zeiler, Dilip Krishnan, Graham W. Taylor, Rob Fergus

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Building robust low and mid-level image representations, beyond edge primitives, is a long-standing goal in vision. Many existing feature detectors spatially pool edge information which destroys cues such as edge intersections, parallelism and symmetry. We present a learning framework where features that capture these mid-level cues spontaneously emerge from image data. Our approach is based on the convolutional decomposition of images under a sparsity constraint and is totally unsupervised. By building a hierarchy of such decompositions we can learn rich feature sets that are a robust image representation for both the analysis and synthesis of images.

Original languageEnglish (US)
Title of host publication2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, CVPR 2010
Pages2528-2535
Number of pages8
DOIs
StatePublished - 2010
Event2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, CVPR 2010 - San Francisco, CA, United States
Duration: Jun 13 2010Jun 18 2010

Other

Other2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, CVPR 2010
CountryUnited States
CitySan Francisco, CA
Period6/13/106/18/10

Fingerprint

Decomposition
Detectors

ASJC Scopus subject areas

  • Software
  • Computer Vision and Pattern Recognition

Cite this

Zeiler, M. D., Krishnan, D., Taylor, G. W., & Fergus, R. (2010). Deconvolutional networks. In 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, CVPR 2010 (pp. 2528-2535). [5539957] https://doi.org/10.1109/CVPR.2010.5539957

Deconvolutional networks. / Zeiler, Matthew D.; Krishnan, Dilip; Taylor, Graham W.; Fergus, Rob.

2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, CVPR 2010. 2010. p. 2528-2535 5539957.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Zeiler, MD, Krishnan, D, Taylor, GW & Fergus, R 2010, Deconvolutional networks. in 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, CVPR 2010., 5539957, pp. 2528-2535, 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, CVPR 2010, San Francisco, CA, United States, 6/13/10. https://doi.org/10.1109/CVPR.2010.5539957
Zeiler MD, Krishnan D, Taylor GW, Fergus R. Deconvolutional networks. In 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, CVPR 2010. 2010. p. 2528-2535. 5539957 https://doi.org/10.1109/CVPR.2010.5539957
Zeiler, Matthew D. ; Krishnan, Dilip ; Taylor, Graham W. ; Fergus, Rob. / Deconvolutional networks. 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, CVPR 2010. 2010. pp. 2528-2535
@inproceedings{0743c2e4d67045e5ae375b81b0450ffc,
title = "Deconvolutional networks",
abstract = "Building robust low and mid-level image representations, beyond edge primitives, is a long-standing goal in vision. Many existing feature detectors spatially pool edge information which destroys cues such as edge intersections, parallelism and symmetry. We present a learning framework where features that capture these mid-level cues spontaneously emerge from image data. Our approach is based on the convolutional decomposition of images under a sparsity constraint and is totally unsupervised. By building a hierarchy of such decompositions we can learn rich feature sets that are a robust image representation for both the analysis and synthesis of images.",
author = "Zeiler, {Matthew D.} and Dilip Krishnan and Taylor, {Graham W.} and Rob Fergus",
year = "2010",
doi = "10.1109/CVPR.2010.5539957",
language = "English (US)",
isbn = "9781424469840",
pages = "2528--2535",
booktitle = "2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, CVPR 2010",

}

TY - GEN

T1 - Deconvolutional networks

AU - Zeiler, Matthew D.

AU - Krishnan, Dilip

AU - Taylor, Graham W.

AU - Fergus, Rob

PY - 2010

Y1 - 2010

N2 - Building robust low and mid-level image representations, beyond edge primitives, is a long-standing goal in vision. Many existing feature detectors spatially pool edge information which destroys cues such as edge intersections, parallelism and symmetry. We present a learning framework where features that capture these mid-level cues spontaneously emerge from image data. Our approach is based on the convolutional decomposition of images under a sparsity constraint and is totally unsupervised. By building a hierarchy of such decompositions we can learn rich feature sets that are a robust image representation for both the analysis and synthesis of images.

AB - Building robust low and mid-level image representations, beyond edge primitives, is a long-standing goal in vision. Many existing feature detectors spatially pool edge information which destroys cues such as edge intersections, parallelism and symmetry. We present a learning framework where features that capture these mid-level cues spontaneously emerge from image data. Our approach is based on the convolutional decomposition of images under a sparsity constraint and is totally unsupervised. By building a hierarchy of such decompositions we can learn rich feature sets that are a robust image representation for both the analysis and synthesis of images.

UR - http://www.scopus.com/inward/record.url?scp=77956001004&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77956001004&partnerID=8YFLogxK

U2 - 10.1109/CVPR.2010.5539957

DO - 10.1109/CVPR.2010.5539957

M3 - Conference contribution

AN - SCOPUS:77956001004

SN - 9781424469840

SP - 2528

EP - 2535

BT - 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, CVPR 2010

ER -