Class-conditional Importance Weighting for Deep Learning with Noisy Labels

Bhalaji Nagarajan, Ricardo Marques, Marcos Mejia, Petia Radeva, Petia Radeva

2022

Abstract

Large-scale accurate labels are very important to the Deep Neural Networks to train them and assure high performance. However, it is very expensive to create a clean dataset since usually it relies on human interaction. To this purpose, the labelling process is made cheap with a trade-off of having noisy labels. Learning with Noisy Labels is an active area of research being at the same time very challenging. The recent advances in Self-supervised learning and robust loss functions have helped in advancing noisy label research. In this paper, we propose a loss correction method that relies on dynamic weights computed based on the model training. We extend the existing Contrast to Divide algorithm coupled with DivideMix using a new class-conditional weighted scheme. We validate the method using the standard noise experiments and achieved encouraging results.

Paper Citation

in Harvard Style

Nagarajan B., Marques R., Mejia M. and Radeva P. (2022). Class-conditional Importance Weighting for Deep Learning with Noisy Labels. In Proceedings of the 17th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2022) - Volume 5: VISAPP; ISBN 978-989-758-555-5, SciTePress, pages 679-686. DOI: 10.5220/0010996400003124

in Bibtex Style

@conference{visapp22,
author={Bhalaji Nagarajan and Ricardo Marques and Marcos Mejia and Petia Radeva},
title={Class-conditional Importance Weighting for Deep Learning with Noisy Labels},
booktitle={Proceedings of the 17th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2022) - Volume 5: VISAPP},
year={2022},
pages={679-686},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0010996400003124},
isbn={978-989-758-555-5},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 17th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2022) - Volume 5: VISAPP
TI - Class-conditional Importance Weighting for Deep Learning with Noisy Labels
SN - 978-989-758-555-5
AU - Nagarajan B.
AU - Marques R.
AU - Mejia M.
AU - Radeva P.
PY - 2022
SP - 679
EP - 686
DO - 10.5220/0010996400003124
PB - SciTePress