Sharmanska_Learning_From_the_CVPR_2016_paper.pdf (2.73 MB)
Learning from the mistakes of others: matching errors in cross dataset learning
conference contribution
posted on 2023-06-09, 00:55 authored by Viktoriia SharmanskaViktoriia Sharmanska, Novi QuadriantoNovi QuadriantoCan we learn about object classes in images by looking at a collection of relevant 3D models? Or if we want to learn about human (inter-)actions in images, can we benefit from videos or abstract illustrations that show these actions? A common aspect of these settings is the availability of additional or privileged data that can be exploited at training time and that will not be available and not of interest at test time. We seek to generalize the learning with privileged information (LUPI) framework, which requires additional information to be defined per image, to the setting where additional information is a data collection about the task of interest. Our framework minimizes the distribution mismatch between errors made in images and in privileged data. The proposed method is tested on four publicly available datasets: Image+ClipArt, Image+3Dobject, and Image+Video. Experimental results reveal that our new LUPI paradigm naturally addresses the cross-dataset learning.
History
Publication status
- Published
File Version
- Accepted version
Journal
Proceedings 29th IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2016); Las Vegas, Nevada; 26 June - 1 July 2016ISSN
1063-6919Publisher
Institute of Electrical and Electronics EngineersExternal DOI
Page range
3967-3975ISBN
9781467388504Department affiliated with
- Informatics Publications
Full text available
- Yes
Peer reviewed?
- Yes
Legacy Posted Date
2016-04-18First Open Access (FOA) Date
2016-09-23First Compliant Deposit (FCD) Date
2016-04-18Usage metrics
Categories
No categories selectedKeywords
Licence
Exports
RefWorks
BibTeX
Ref. manager
Endnote
DataCite
NLM
DC