DropFilter: Dropout for Convolutions

TL;DR

DropFilter is a novel dropout technique for convolutional layers that suppresses filter outputs to reduce overfitting and improve neural network performance on image classification tasks.

Contribution

This paper introduces DropFilter, a new dropout method specifically designed for convolutional layers, addressing limitations of traditional dropout in CNNs.

Findings

Significantly improves CNN performance on CIFAR and ImageNet datasets.

Effectively reduces overfitting in convolutional neural networks.

Addresses co-adaptation issues among filters in convolutional layers.

Abstract

Using a large number of parameters , deep neural networks have achieved remarkable performance on computer vison and natural language processing tasks. However the networks usually suffer from overfitting by using too much parameters. Dropout is a widely use method to deal with overfitting. Although dropout can significantly regularize densely connected layers in neural networks, it leads to suboptimal results when using for convolutional layers. To track this problem, we propose DropFilter, a new dropout method for convolutional layers. DropFilter randomly suppresses the outputs of some filters. Because it is observed that co-adaptions are more likely to occurs inter filters rather than intra filters in convolutional layers. Using DropFilter, we remarkably improve the performance of convolutional networks on CIFAR and ImageNet.