Articles liés à AUTOENCODER NEURAL NETWORKS: A Performance Study Based...

AUTOENCODER NEURAL NETWORKS: A Performance Study Based on Image Reconstruction, Recognition and Compression - Couverture souple

Tan, Chun Chet

 
9783838309460: AUTOENCODER NEURAL NETWORKS: A Performance Study Based on Image Reconstruction, Recognition and Compression
Couverture souple
ISBN 10 : 3838309464 ISBN 13 : 9783838309460
Editeur : LAP Lambert Academic Publishing, 2009

Synopsis

Autoencoders are feedforward neural networks which can have more than one hidden layer. These networks attempt to reconstruct the input data at the output layer. Since the size of the hidden layer in the autoencoders is smaller than the size of the input data, the dimensionality of input data is reduced to a smaller-dimensional code space at the hidden layer. However, training a multilayer autoencoder is tedious. This is due to the fact that the weights at deep hidden layers are hardly optimized. The research work has focused on the characteristics, training and performance evaluation of autoencoders. The concepts of stacking and Restricted Boltzmann Machine have also been discussed in detail. Two datasets, namely ORL face dataset and MNIST handwritten digit dataset have been employed in these experiments. The performances of the autoencoders have also been compared with that of PCA. It has been shown that the autoencoders can also be used for image compression. The compression efficiency has been studied using DDSM dataset (mammogram dataset). Since image patches were used for training, it was possible to compress and decompress mammograms of different sizes.

Les informations fournies dans la section « Synopsis » peuvent faire référence à une autre édition de ce titre.

Présentation de l'éditeur

Autoencoders are feedforward neural networks which can have more than one hidden layer. These networks attempt to reconstruct the input data at the output layer. Since the size of the hidden layer in the autoencoders is smaller than the size of the input data, the dimensionality of input data is reduced to a smaller-dimensional code space at the hidden layer. However, training a multilayer autoencoder is tedious. This is due to the fact that the weights at deep hidden layers are hardly optimized. The research work has focused on the characteristics, training and performance evaluation of autoencoders. The concepts of stacking and Restricted Boltzmann Machine have also been discussed in detail. Two datasets, namely ORL face dataset and MNIST handwritten digit dataset have been employed in these experiments. The performances of the autoencoders have also been compared with that of PCA. It has been shown that the autoencoders can also be used for image compression. The compression efficiency has been studied using DDSM dataset (mammogram dataset). Since image patches were used for training, it was possible to compress and decompress mammograms of different sizes.

Biographie de l'auteur

C. C. Tan: received B.Sc. degree from University of Malaya, and M.IT degree from Multimedia University, Malaysia. He is a lecturer in Faculty of I.T., Multimedia University. Dr. C. Eswaran: received B.Tech, M.Tech and Ph.D degrees from the Indian Institute of Technolgy Madras, India. He is a Professor in Faculty of I.T., Multimedia University.

Les informations fournies dans la section « A propos du livre » peuvent faire référence à une autre édition de ce titre.

Éditeur
LAP Lambert Academic Publishing
Date d'édition
2009
Langue
anglais
ISBN 10 
3838309464
ISBN 13 
9783838309460
Reliure
Broché
Nombre de pages
96
Coordonnées du fabricant
non disponible
Personne responsable
KreGa Onlinehandel GmbH
4950000000000
info@krega-onlinehandel.de

Birkenweg 31
Brakel
33034
Allemagne