Affordable Access

ReNet: A Recurrent Neural Network Based Alternative to Convolutional Networks

Authors
  • Visin, Francesco
  • Kastner, Kyle
  • Cho, Kyunghyun
  • Matteucci, Matteo
  • Courville, Aaron
  • Bengio, Yoshua
Type
Preprint
Publication Date
Jul 23, 2015
Submission Date
May 03, 2015
Identifiers
arXiv ID: 1505.00393
Source
arXiv
License
Yellow
External links

Abstract

In this paper, we propose a deep neural network architecture for object recognition based on recurrent neural networks. The proposed network, called ReNet, replaces the ubiquitous convolution+pooling layer of the deep convolutional neural network with four recurrent neural networks that sweep horizontally and vertically in both directions across the image. We evaluate the proposed ReNet on three widely-used benchmark datasets; MNIST, CIFAR-10 and SVHN. The result suggests that ReNet is a viable alternative to the deep convolutional neural network, and that further investigation is needed.

Report this publication

Statistics

Seen <100 times