Toward Robustness in Multi-Label Classification: A Data Augmentation Strategy against Imbalance and Noise
Proceedings of the AAAI Conference on Artificial Intelligence · 24.03.2024 · 38 (19) · 21592-21601
DOI: 10.1609/aaai.v38i19.30157 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 24.03.2024. Verfasst von Hwanjun Song und 2 weiteren Autor:innen. Die Arbeit wurde bisher 28-mal zitiert (Crossref).
Abstract (Originalsprache)
Multi-label classification poses challenges due to imbalanced and noisy labels in training data. In this paper, we propose a unified data augmentation method, named BalanceMix, to address these challenges. Our approach includes two samplers for imbalanced labels, generating minority-augmented instances with high diversity. It also refines multi-labels at the label-wise granularity, categorizing noisy labels as clean, re-labeled, or ambiguous for robust optimization. Extensive experiments on three benchmark datasets demonstrate that BalanceMix outperforms existing state-of-the-art methods. We release the code at https://github.com/DISL-Lab/BalanceMix.
Zitieren
APA
Song, H., Kim, M., & Lee, J. (2024). Toward Robustness in Multi-Label Classification: A Data Augmentation Strategy against Imbalance and Noise. Proceedings of the AAAI Conference on Artificial Intelligence, 38(19), 21592-21601. https://doi.org/10.1609/aaai.v38i19.30157
Vancouver
Song H, Kim M, Lee J. Toward Robustness in Multi-Label Classification: A Data Augmentation Strategy against Imbalance and Noise. Proceedings of the AAAI Conference on Artificial Intelligence. 2024;38(19):21592-21601. doi:10.1609/aaai.v38i19.30157
BibTeX
@article{Song2024,
author = {Song, Hwanjun and Kim, Minseok and Lee, Jae-Gil},
title = {Toward Robustness in Multi-Label Classification: A Data Augmentation Strategy against Imbalance and Noise},
journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
year = {2024},
volume = {38},
number = {19},
pages = {21592--21601},
doi = {10.1609/aaai.v38i19.30157}
}
EUVIMED