Gomes, Ryan and Welinder, Peter and Krause, Andreas and Perona, Pietro (2011) Crowdclustering. Computation & Neural Systems Technical Report, CNS-TR. California Institute of Technology , Pasadena, CA. (Submitted) http://resolver.caltech.edu/CaltechAUTHORS:20111027-114015448
This is the latest version of this item.
|
PDF
- Published Version
See Usage Policy. 7Mb |
Use this Persistent URL to link to this item: http://resolver.caltech.edu/CaltechAUTHORS:20111027-114015448
Abstract
Is it possible to crowdsource categorization? Amongst the challenges: (a) each worker has only a partial view of the data, (b) different workers may have different clustering criteria and may produce different numbers of categories, (c) the underlying category structure may be hierarchical. We propose a Bayesian model of how workers may approach clustering and show how one may infer clusters / categories, as well as worker parameters, using this model. Our experiments, carried out on large collections of images, suggest that Bayesian crowdclustering works well and may be superior to single-expert annotations.
| Item Type: | Report or Paper (Technical Report) | ||||||
|---|---|---|---|---|---|---|---|
| Additional Information: | This work was supported by ONR MURI grant N00014-06-1-0734 and NSF grant IIS-0953413. | ||||||
| Group: | Computation & Neural Systems Technical Reports | ||||||
| Funders: |
| ||||||
| Record Number: | CaltechAUTHORS:20111027-114015448 | ||||||
| Persistent URL: | http://resolver.caltech.edu/CaltechAUTHORS:20111027-114015448 | ||||||
| Usage Policy: | No commercial reproduction, distribution, display or performance rights in this work are provided. | ||||||
| ID Code: | 27473 | ||||||
| Collection: | CaltechAUTHORS | ||||||
| Deposited By: | Ryan Gomes | ||||||
| Deposited On: | 09 Nov 2011 00:01 | ||||||
| Last Modified: | 26 Dec 2012 14:20 |
Available Versions of this Item
-
Crowdclustering. (deposited 29 Jun 2011 21:29)
- Crowdclustering. (deposited 09 Nov 2011 00:01) [Currently Displayed]
Repository Staff Only: item control page


