Robust Long-Tailed Image Classification via Adversarial Feature Re-Calibration

Jinghao Zhang, Zhenhua Feng, Yaochu Jin

2024

Abstract

Long-tailed data distribution is a common issue in many practical learning-based approaches, causing Deep Neural Networks (DNNs) to under-fit minority classes. Although this biased problem has been extensively studied by the research community, the existing approaches mainly focus on the class-wise (inter-class) imbalance problem. In contrast, this paper considers both inter-class and intra-class data imbalance problems for network training. To this end, we present Adversarial Feature Re-calibration (AFR), a method that improves the standard accuracy of a trained deep network by adding adversarial perturbations to the majority samples of each class. To be specific, an adversarial attack model is fine-tuned to perturb the majority samples by injecting the features from their corresponding intra-class long-tailed minority samples. This procedure makes the dataset more evenly distributed from both the inter- and intra-class perspectives, thus encouraging DNNs to learn better representations. The experimental results obtained on CIFAR-100-LT demonstrate the effectiveness and superiority of the proposed AFR method over the state-of-the-art long-tailed learning methods.

Paper Citation

in Harvard Style

Zhang J., Feng Z. and Jin Y. (2024). Robust Long-Tailed Image Classification via Adversarial Feature Re-Calibration. In Proceedings of the 19th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 2: VISAPP; ISBN 978-989-758-679-8, SciTePress, pages 213-220. DOI: 10.5220/0012432000003660

in Bibtex Style

@conference{visapp24,
author={Jinghao Zhang and Zhenhua Feng and Yaochu Jin},
title={Robust Long-Tailed Image Classification via Adversarial Feature Re-Calibration},
booktitle={Proceedings of the 19th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 2: VISAPP},
year={2024},
pages={213-220},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0012432000003660},
isbn={978-989-758-679-8},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 19th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 2: VISAPP
TI - Robust Long-Tailed Image Classification via Adversarial Feature Re-Calibration
SN - 978-989-758-679-8
AU - Zhang J.
AU - Feng Z.
AU - Jin Y.
PY - 2024
SP - 213
EP - 220
DO - 10.5220/0012432000003660
PB - SciTePress