期刊
IEEE ACCESS
卷 7, 期 -, 页码 186126-186136出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/ACCESS.2019.2960566
关键词
Computer vision; image classification; neural networks
资金
- JST CREST [JPMJCR16E1]
- JSPS KAKENHI [JP25240028, JP17H01803, JP18J15255, JP18K19785]
- JST AIP PRISM [J18ZZ00418]
- Artificial Intelligence Research Promotion Foundation
- AWS Cloud Credits for Research Program
Overfitting is a crucial problem in deep neural networks, even in the latest network architectures. In this paper, to relieve the overfitting effect of ResNet and its improvements (i.e., Wide ResNet, PyramidNet, and ResNeXt), we propose a new regularization method called ShakeDrop regularization. ShakeDrop is inspired by Shake-Shake, which is an effective regularization method, but can be applied to ResNeXt only. ShakeDrop is more effective than Shake-Shake and can be applied not only to ResNeXt but also ResNet, Wide ResNet, and PyramidNet. An important key is to achieve stability of training. Because effective regularization often causes unstable training, we introduce a training stabilizer, which is an unusual use of an existing regularizer. Through experiments under various conditions, we demonstrate the conditions under which ShakeDrop works well.
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