ShakeDrop regularization
ShakeDrop regularization
Yoshihiro Yamada,M. Iwamura,K. Kise
2018 · DBLP: journals/corr/abs-1802-02375
International Conference on Learning Representations · 70 Citations
TLDR
ShakeDrop is inspired by Shake-Shake regularization that decreases error rates by disturbing learning and can be applied to not only ResNeXt but also ResNet, Wide ResNet and PyramidNet in a memory efficient way.