English
 
Help Privacy Policy Disclaimer
  Advanced SearchBrowse

Item

ITEM ACTIONSEXPORT
 
 
DownloadE-Mail
 DetailsSummary
  Preserving Modes and Messages via Diverse Particle Selection

Pacheco, J., Zuffi, S., Black, M. J., & Sudderth, E. (2014). Preserving Modes and Messages via Diverse Particle Selection. In E. P. Xing, & T. Jebara (Eds.), Proceedings of the 31st International Conference on Machine Learning (ICML 2014) (pp. 1152-1160). Brookline, MA: Microtome Publishing. Retrieved from http://jmlr.csail.mit.edu/proceedings/papers/v32/pacheco14.pdf.

Item is

 

show all hide all

Basic

show hide
Item Permalink: https://hdl.handle.net/11858/00-001M-0000-0024-DD2B-6 Version Permalink: https://hdl.handle.net/11858/00-001M-0000-0025-B84D-0
Genre: Conference Paper

Files

show Files

Locators

show

Creators

show
hide
 Creators:
Pacheco, Jason, Author
Zuffi, Sylvia1, Author           
Black, Michael J.1, Author           
Sudderth, Erik, Author
Affiliations:
1Dept. Perceiving Systems, Max Planck Institute for Intelligent Systems, Max Planck Society, ou_1497642              

Content

show
hide
Free keywords: Abt. Black
 Abstract: In applications of graphical models arising in domains such as computer vision and signal processing, we often seek the most likely configurations of high-dimensional, continuous variables. We develop a particle-based max-product algorithm which maintains a diverse set of posterior mode hypotheses, and is robust to initialization. At each iteration, the set of hypotheses at each node is augmented via stochastic proposals, and then reduced via an efficient selection algorithm. The integer program underlying our optimization-based particle selection minimizes errors in subsequent max-product message updates. This objective automatically encourages diversity in the maintained hypotheses, without requiring tuning of application-specific distances among hypotheses. By avoiding the stochastic resampling steps underlying particle sum-product algorithms, we also avoid common degeneracies where particles collapse onto a single hypothesis. Our approach significantly outperforms previous particle-based algorithms in experiments focusing on the estimation of human pose from single images.

Details

show
hide
Language(s):
 Dates: 2014-06
 Publication Status: Published online
 Pages: -
 Publishing info: -
 Table of Contents: -
 Rev. Type: -
 Identifiers: BibTex Citekey: Pacheco:ICML:2014
URI: http://jmlr.csail.mit.edu/proceedings/papers/v32/pacheco14.pdf
 Degree: -

Event

show
hide
Title: 31st International Conference on Machine Learning (ICML 2014)
Place of Event: Beijing, China
Start-/End Date: 2014-06-21 - 2014-06-26

Legal Case

show

Project information

show

Source 1

show
hide
Title: Proceedings of the 31st International Conference on Machine Learning (ICML 2014)
Source Genre: Proceedings
 Creator(s):
Xing, E. P., Editor
Jebara, T., Editor
Affiliations:
-
Publ. Info: Brookline, MA : Microtome Publishing
Pages: - Volume / Issue: - Sequence Number: - Start / End Page: 1152 - 1160 Identifier: -

Source 2

show
hide
Title: JMLR: Workshop & Conference Proceedings
  Abbreviation : JMLR: W&CP
Source Genre: Series
 Creator(s):
Affiliations:
Publ. Info: -
Pages: - Volume / Issue: 32 Sequence Number: - Start / End Page: - Identifier: ISSN: 1938-7228