Modified kNN Classifier in the Output Vector Space for Robust Performance Against Adversarial Attack

C. Lee, D. Seok, D. Shim, R. Park

2023

Abstract

Although CNN-based classifiers have been successfully applied to many pattern classification problems, they suffer from adversarial attacks. Slightly modified images can be classified as completely different classes. It has been reported that CNN-based classifiers tend to construct decision boundaries close to training samples. In order to mitigate this problem, we applied modified kNN classifiers in the output vector space of CNN-based classifiers. Experimental results show that the proposed method noticeably reduced the classification error caused by adversarial attacks.

Paper Citation

in Harvard Style

Lee C., Seok D., Shim D. and Park R. (2023). Modified kNN Classifier in the Output Vector Space for Robust Performance Against Adversarial Attack. In Proceedings of the 12th International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM, ISBN 978-989-758-626-2, pages 443-449. DOI: 10.5220/0011735800003411

in Bibtex Style

@conference{icpram23,
author={C. Lee and D. Seok and D. Shim and R. Park},
title={Modified kNN Classifier in the Output Vector Space for Robust Performance Against Adversarial Attack},
booktitle={Proceedings of the 12th International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM,},
year={2023},
pages={443-449},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0011735800003411},
isbn={978-989-758-626-2},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 12th International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM,
TI - Modified kNN Classifier in the Output Vector Space for Robust Performance Against Adversarial Attack
SN - 978-989-758-626-2
AU - Lee C.
AU - Seok D.
AU - Shim D.
AU - Park R.
PY - 2023
SP - 443
EP - 449
DO - 10.5220/0011735800003411