Stacked Convolutional Auto-Encoders for Hierarchical Feature Extraction

被引：1497

作者：

Masci, Jonathan ^{[1
]}

Meier, Ueli ^{[1
]}

Ciresan, Dan ^{[1
]}

Schmidhuber, Juergen ^{[1
]}

机构：

[1] IDSIA, Lugano, Switzerland

来源：

ARTIFICIAL NEURAL NETWORKS AND MACHINE LEARNING - ICANN 2011, PT I | 2011年 / 6791卷

关键词：

convolutional neural network; auto-encoder; unsupervised learning; classification;

D O I：

10.1007/978-3-642-21735-7_7

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

We present a novel convolutional auto-encoder (CAE) for unsupervised feature learning. A stack of CAEs forms a convolutional neural network (CNN). Each CAE is trained using conventional on-line gradient descent without additional regularization terms. A max-pooling layer is essential to learn biologically plausible features consistent with those found by previous approaches. Initializing a CNN with filters of a trained CAE stack yields superior performance on a digit (MNIST) and an object recognition (CIFAR10) benchmark.

引用

页码：52 / 59

页数：8