Stochastic Optimization of Plain Convolutional Neural Networks with Simple methods

TL;DR

This paper demonstrates that combining simple regularization techniques like data augmentation, dropout, and early stopping on plain CNNs can significantly improve performance across multiple datasets, achieving state-of-the-art results.

Contribution

The study shows that straightforward regularization methods, when combined, effectively enhance CNN generalization without complex architectures.

Findings

Achieved state-of-the-art results on MNIST, SVHN, STL10 datasets.

High accuracy obtained on CIFAR10 and CIFAR100 datasets.

Simple methods outperform some complex models in certain benchmarks.

Abstract

Convolutional neural networks have been achieving the best possible accuracies in many visual pattern classification problems. However, due to the model capacity required to capture such representations, they are often oversensitive to overfitting and therefore require proper regularization to generalize well. In this paper, we present a combination of regularization techniques which work together to get better performance, we built plain CNNs, and then we used data augmentation, dropout and customized early stopping function, we tested and evaluated these techniques by applying models on five famous datasets, MNIST, CIFAR10, CIFAR100, SVHN, STL10, and we achieved three state-of-the-art-of (MNIST, SVHN, STL10) and very high-Accuracy on the other two datasets.