Comparison of Regularization Methods for ImageNet Classification with Deep Convolutional Neural Networks

Evgeny A. Smirnov, Denis M. Timoshenko, Serge N. Andrianov

Research output

Abstract

Large and Deep Convolutional Neural Networks achieve good results in image classification tasks, but they need methods to prevent overfitting. In this paper we compare performance of different regularization techniques on ImageNet Large Scale Visual Recognition Challenge 2013. We show empirically that Dropout works better than DropConnect on ImageNet dataset.
Original languageEnglish
Pages (from-to)89-94
JournalAASRI Procedia
Volume6
Publication statusPublished - 2014

Fingerprint Dive into the research topics of 'Comparison of Regularization Methods for ImageNet Classification with Deep Convolutional Neural Networks'. Together they form a unique fingerprint.

Cite this