期刊
MEDICAL IMAGE COMPUTING AND COMPUTER ASSISTED INTERVENTION - MICCAI 2021, PT V
卷 12905, 期 -, 页码 323-333出版社
SPRINGER INTERNATIONAL PUBLISHING AG
DOI: 10.1007/978-3-030-87240-3_31
关键词
Imbalanced learning; Long-tail image classification
类别
资金
- Marie SklodowskaCurie Global Fellowship [892297]
- Australian Research Council [DP180103232, FT190100525]
- Marie Curie Actions (MSCA) [892297] Funding Source: Marie Curie Actions (MSCA)
In this paper, a novel mechanism called Balanced-MixUp is proposed to address highly imbalanced medical image classification problems, it improves the balance of training data by simultaneously performing regular and balanced sampling, and experiments show promising results compared to traditional approaches.
Highly imbalanced datasets are ubiquitous in medical image classification problems. In such problems, it is often the case that rare classes associated to less prevalent diseases are severely under-represented in labeled databases, typically resulting in poor performance of machine learning algorithms due to overfitting in the learning process. In this paper, we propose a novel mechanism for sampling training data based on the popular MixUp regularization technique, which we refer to as Balanced-MixUp. In short, Balanced-MixUp simultaneously performs regular (i.e., instance-based) and balanced (i.e., class-based) sampling of the training data. The resulting two sets of samples are then mixed-up to create a more balanced training distribution from which a neural network can effectively learn without incurring in heavily under-fitting the minority classes. We experiment with a highly imbalanced dataset of retinal images (55K samples, 5 classes) and a long-tail dataset of gastro-intestinal video frames (10K images, 23 classes), using two CNNs of varying representation capabilities. Experimental results demonstrate that applying Balanced-MixUp outperforms other conventional sampling schemes and loss functions specifically designed to deal with Unbalanced data. Code is released at https://github.com/agaldran/balanced_mixup
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据