Undoing the Damage of Dataset Bias
Name
eccv2012_khosla.pdf
Description
Accepted version
Size
3.85 MB
Format
Adobe PDF
Checksum (MD5)
7433a7e6641f5c51b8f8d0f846cace66
Author(s)
Khosla, Aditya
Date Issued
September 2012
Journal
European Conference on Computer Vision
Publisher
Springer Nature
Citation
Kholsa, Aditya et al. "Undoing the Damage of Dataset Bias." European Conference on Computer Vision, September 2012, Munich, Germany, Springer Nature, 2012 © 2012 Springer-Verlag
Version
Author's final manuscript
Abstract
The presence of bias in existing object recognition datasets is now well-known in the computer vision community. While it remains in question whether creating an unbiased dataset is possible given limited resources, in this work we propose a discriminative framework that directly exploits dataset bias during training. In particular, our model learns two sets of weights: (1) bias vectors associated with each individual dataset, and (2) visual world weights that are common to all datasets, which are learned by undoing the associated bias from each dataset. The visual world weights are expected to be our best possible approximation to the object model trained on an unbiased dataset, and thus tend to have good generalization ability. We demonstrate the effectiveness of our model by applying the learned weights to a novel, unseen dataset, and report superior results for both classification and detection tasks compared to a classical SVM that does not account for the presence of bias. Overall, we find that it is beneficial to explicitly account for bias when combining multiple datasets. Keywords: Target Domain; Domain Adaptation; Transfer Learning; Visual World; Spatial Pyramid
MIT Department
Massachusetts Institute of Technology. Media Laboratory
Terms of Use
Creative Commons Attribution-Noncommercial-Share Alike
Persistent DSpace Link
DOI of Published Version
https://doi.org/10.1007/978-3-642-33718-5_12
