ShakeDrop Regularization for Deep Residual Learning

Yoshihiro Yamada, Masakazu Iwamura, Takuya Akiba, Koichi Kise
2/7/2018
cs.CV

Official code: imenurok/ShakeDrop (21 stars, last commit 2018-02-08).

Abstract

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.

View on arXivView PDF

Code Implementations(1)

imenurok/ShakeDropOfficial100%

Torch implementation of the paper "ShakeDrop regularization" (https://arxiv.org/abs/1802.02375).

217LuaFeb 8, 20188 years ago

Similar Papers

Drop-Activation: Implicit Parameter Reduction and Harmonic Regularization

89% match1 repo102018

R-Drop: Regularized Dropout for Neural Networks

85% match1 repo8792021

DELTA: DEep Learning Transfer using Feature Map with Attention for Convolutional Networks

85% match1 repo662019

Counterbalancing Teacher: Regularizing Batch Normalized Models for Robustness

85% match2022

Learning from Teaching Regularization: Generalizable Correlations Should be Easy to Imitate

85% match1 repo332024

Recurrent Neural Network Regularization

84% match1 repo6822014

Cite this paper

@article{yamada2018shakedrop,
  title  = {ShakeDrop Regularization for Deep Residual Learning},
  author = {Yoshihiro Yamada and Masakazu Iwamura and Takuya Akiba and Koichi Kise},
  year   = {2018},
  eprint = {1802.02375},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1802.02375}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/1802.02375.svg)](https://codeofpaper.com/paper/1802.02375)

Discussion