Revisiting Uncertainty in Graph Cut Solutions

DSpace/Manakin Repository

Revisiting Uncertainty in Graph Cut Solutions

Show simple item record

dc.contributor.author Adams, Ryan Prescott
dc.contributor.author Tarlow, Daniel
dc.date.accessioned 2012-05-10T13:16:26Z
dc.date.issued 2012
dc.identifier.citation Tarlow, Daniel and Ryan P Adams. Forthcoming. Revisiting uncertainty in graph cut solutions. Proceedings of the IEEE Conference on Computer Vision Pattern Recognition (CVPR): June 16-21, 2012, Providence, Rhode Island. en_US
dc.identifier.uri http://nrs.harvard.edu/urn-3:HUL.InstRepos:8712188
dc.description.abstract Graph cuts is a popular algorithm for finding the MAP assignment of many large-scale graphical models that are common in computer vision. While graph cuts is powerful, it does not provide information about the marginal probabilities associated with the solution it finds. To assess uncertainty, we are forced to fall back on less efficient and inexact inference algorithms such as loopy belief propagation, or use less principled surrogate representations of uncertainty such as the min-marginal approach of Kohli & Torr. In this work, we give new justification for using min-marginals to compute the uncertainty in conditional random fields, framing the min-marginal outputs as exact marginals under a specially-chosen generative probabilistic model. We leverage this view to learn properly calibrated marginal probabilities as the result of straightforward maximization of the training likelihood, showing that the necessary subgradients can be computed efficiently using dynamic graph cut operations. We also show how this approach can be extended to compute multi-label marginal distributions, where again dynamic graph cuts enable efficient marginal inference and maximum likelihood learning. We demonstrate empirically that — after proper training — uncertainties based on min-marginals provide better- calibrated probabilities than baselines and that these distributions can be exploited in a decision-theoretic way for improved segmentation in low-level vision. en_US
dc.description.sponsorship Engineering and Applied Sciences en_US
dc.language.iso en_US en_US
dc.publisher Institute of Electrical and Electronics Engineers en_US
dash.license OAP
dc.title Revisiting Uncertainty in Graph Cut Solutions en_US
dc.type Journal Article en_US
dc.description.version Accepted Manuscript en_US
dc.relation.journal Proceedings of the IEEE Conference on Computer Vision Pattern Recognition (CVPR) en_US
dash.depositing.author Adams, Ryan Prescott
dc.date.available 2012-05-10T13:16:26Z

Files in this item

Files Size Format View
tarlow-revisiting-cvpr-2012.pdf 5.133Mb PDF View/Open

This item appears in the following Collection(s)

  • FAS Scholarly Articles [7501]
    Peer reviewed scholarly articles from the Faculty of Arts and Sciences of Harvard University

Show simple item record

 
 

Search DASH


Advanced Search
 
 

Submitters