Please use this identifier to cite or link to this item: http://hdl.handle.net/10553/55092
Title: Morphological convolutional neural network architecture for digit recognition
Authors: Mellouli, Dorra
Hamdani, Tarek M.
Sanchez-Medina, Javier J. 
Ayed, Mounir Ben
Alimi, Adel M.
UNESCO Clasification: 120304 Inteligencia artificial
Keywords: Convolutional neural network (CNN)
Deep neural networks (DNNs)
Image recognition
Interpretability
Morphological CNN (Morph-CNN)
Morphological operators
Issue Date: 2019
Publisher: 2162-237X
Journal: IEEE Transactions on Neural Networks and Learning Systems 
Abstract: Deep neural networks have proved promising results in many applications and fields, but they are still assimilated to a black box. Thus, it is very useful to introduce interpretability aspects to prevent the blind application of deep networks. This paper proposed an interpretable morphological convolutional neural network called Morph-CNN for pattern recognition, where morphological operations were incorporated using counter-harmonic mean into the convolutional layer in order to generate enhanced feature maps. Morph-CNN was extensively evaluated on MNIST and SVHN benchmarks for digit recognition. The different tested configurations showed that Morph-CNN outperforms the existing methods.
URI: http://hdl.handle.net/10553/55092
ISSN: 2162-237X
DOI: 10.1109/TNNLS.2018.2890334
Source: IEEE Transactions on Neural Networks and Learning Systems [ISSN 2162-237X], v. 30 (9), p. 2876 - 2885
Appears in Collections:Artículos
Show full item record

SCOPUSTM   
Citations

1
checked on May 31, 2020

WEB OF SCIENCETM
Citations

1
checked on May 31, 2020

Page view(s)

47
checked on May 30, 2020

Google ScholarTM

Check

Altmetric


Share



Export metadata



Items in accedaCRIS are protected by copyright, with all rights reserved, unless otherwise indicated.