A Full Stage Data Augmentation Method in Deep Convolutional Neural Network for Natural Image Classification

Table 2

Comparison with state-of-the-art algorithms on CIFAR-10.

Algorithms	CIFAR-10 (%)
Dropout [41]	84.40
Probout [42]	88.65
NIN + dropout [43]	89.59
Maxout + dropout [44]	88.32
Stochastic pooling [45]	84.86
Probabilistic weighted pooling [46]	88.71
Our method	93.41