Convolutional neural networks with fused layers applied to face recognition
In this paper, we propose an e®ective convolutional neural network (CNN) model to the problem of face recognition. The proposed CNN architecture applies fused convolution/ subsampling layers that result in a simpler model with fewer network parameters; that is, a smaller number of neurons, trainabl...
Saved in:
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
World Scientific Publishing
2015
|
Online Access: | http://eprints.utem.edu.my/id/eprint/18946/2/CNNs%20with%20fused%20layers%20applied%20to%20face%20recognition.pdf http://eprints.utem.edu.my/id/eprint/18946/ http://www.worldscientific.com/action/showMultipleAbstracts?mailPageTitle=Search&href=%2Faction%2FdoSearch%3FpubType%3D%26AllField%3DConvolutional%2BNeural%2BNetworks%2BWith%2BFused%2BConvolution%252FSubsampling%2BLayers%2BApplied%2BTo%2BFace%2BRecognitio |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
id |
my.utem.eprints.18946 |
---|---|
record_format |
eprints |
spelling |
my.utem.eprints.189462023-07-04T15:23:15Z http://eprints.utem.edu.my/id/eprint/18946/ Convolutional neural networks with fused layers applied to face recognition Ahmad Radzi, Syafeeza Hani, Mohamed Khalil Liew, Shan Sung Bakhteri, Rabia In this paper, we propose an e®ective convolutional neural network (CNN) model to the problem of face recognition. The proposed CNN architecture applies fused convolution/ subsampling layers that result in a simpler model with fewer network parameters; that is, a smaller number of neurons, trainable parameters, and connections. In addition, it does not require any complex or costly image preprocessing steps that are typical in existing face recognizer systems. In this work, we enhance the stochastic diagonal Levenberg–Marquardt algorithm, a second-order back-propagation algorithm to obtain faster network convergence and better generalization ability. Experimental work completed on the ORL database shows that a recognition accuracy of 100% is achieved, with the network converging within 15 epochs. The average processing time of the proposed CNN face recognition solution, executed on a 2.5 GHz Intel i5 quad-core processor, is 3 s per epoch, with a recognition speed of less than 0.003 s. These results show that the proposed CNN model is a computationally efficient architecture that exhibits faster processing and learning times, and also produces higher recognition accuracy, outperforming other existing work on face recognizers based on neural networks World Scientific Publishing 2015 Article PeerReviewed text en http://eprints.utem.edu.my/id/eprint/18946/2/CNNs%20with%20fused%20layers%20applied%20to%20face%20recognition.pdf Ahmad Radzi, Syafeeza and Hani, Mohamed Khalil and Liew, Shan Sung and Bakhteri, Rabia (2015) Convolutional neural networks with fused layers applied to face recognition. International Journal of Computational Intelligence and Applications, 14 (3). pp. 1-19. ISSN 1469-0268 http://www.worldscientific.com/action/showMultipleAbstracts?mailPageTitle=Search&href=%2Faction%2FdoSearch%3FpubType%3D%26AllField%3DConvolutional%2BNeural%2BNetworks%2BWith%2BFused%2BConvolution%252FSubsampling%2BLayers%2BApplied%2BTo%2BFace%2BRecognitio 10.1142/S1469026815500145 |
institution |
Universiti Teknikal Malaysia Melaka |
building |
UTEM Library |
collection |
Institutional Repository |
continent |
Asia |
country |
Malaysia |
content_provider |
Universiti Teknikal Malaysia Melaka |
content_source |
UTEM Institutional Repository |
url_provider |
http://eprints.utem.edu.my/ |
language |
English |
description |
In this paper, we propose an e®ective convolutional neural network (CNN) model to the problem of face recognition. The proposed CNN architecture applies fused convolution/
subsampling layers that result in a simpler model with fewer network parameters; that is, a smaller number of neurons, trainable parameters, and connections. In addition, it does not
require any complex or costly image preprocessing steps that are typical in existing face recognizer systems. In this work, we enhance the stochastic diagonal Levenberg–Marquardt algorithm, a second-order back-propagation algorithm to obtain faster network convergence and better generalization ability. Experimental work completed on the ORL database shows that a recognition accuracy of 100% is achieved, with the network converging within 15 epochs. The average processing time of the proposed CNN face recognition solution, executed on a 2.5 GHz Intel i5 quad-core processor, is 3 s per epoch, with a recognition speed of less than
0.003 s. These results show that the proposed CNN model is a computationally efficient architecture that exhibits faster processing and learning times, and also produces higher recognition accuracy, outperforming other existing work on face recognizers based on neural networks |
format |
Article |
author |
Ahmad Radzi, Syafeeza Hani, Mohamed Khalil Liew, Shan Sung Bakhteri, Rabia |
spellingShingle |
Ahmad Radzi, Syafeeza Hani, Mohamed Khalil Liew, Shan Sung Bakhteri, Rabia Convolutional neural networks with fused layers applied to face recognition |
author_facet |
Ahmad Radzi, Syafeeza Hani, Mohamed Khalil Liew, Shan Sung Bakhteri, Rabia |
author_sort |
Ahmad Radzi, Syafeeza |
title |
Convolutional neural networks with fused layers applied to face recognition |
title_short |
Convolutional neural networks with fused layers applied to face recognition |
title_full |
Convolutional neural networks with fused layers applied to face recognition |
title_fullStr |
Convolutional neural networks with fused layers applied to face recognition |
title_full_unstemmed |
Convolutional neural networks with fused layers applied to face recognition |
title_sort |
convolutional neural networks with fused layers applied to face recognition |
publisher |
World Scientific Publishing |
publishDate |
2015 |
url |
http://eprints.utem.edu.my/id/eprint/18946/2/CNNs%20with%20fused%20layers%20applied%20to%20face%20recognition.pdf http://eprints.utem.edu.my/id/eprint/18946/ http://www.worldscientific.com/action/showMultipleAbstracts?mailPageTitle=Search&href=%2Faction%2FdoSearch%3FpubType%3D%26AllField%3DConvolutional%2BNeural%2BNetworks%2BWith%2BFused%2BConvolution%252FSubsampling%2BLayers%2BApplied%2BTo%2BFace%2BRecognitio |
_version_ |
1770555158074228736 |
score |
13.160551 |