Gender classication: a convolutional neural network approach
An approach using a convolutional neural network (CNN) is proposed for real-time gender classication based on facial images. The proposed CNN architecture exhibits a much reduced design complexity when compared with other CNN solutions applied in pattern recognition. The number of processing layers...
Saved in:
Main Authors: | , , , |
---|---|
Format: | Article |
Published: |
TUBITAK
2016
|
Subjects: | |
Online Access: | http://eprints.utm.my/id/eprint/68390/ http://dx.doi.org/10.3906/elk-1311-58 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Universiti Teknologi Malaysia |
id |
my.utm.68390 |
---|---|
record_format |
eprints |
spelling |
my.utm.683902017-11-20T08:52:08Z http://eprints.utm.my/id/eprint/68390/ Gender classication: a convolutional neural network approach Liew, Shan Sung Mohd. Hani, Mohamed Khalil Ahmad Radzi, Syafeeza Bakhteri, Rabia TK Electrical engineering. Electronics Nuclear engineering An approach using a convolutional neural network (CNN) is proposed for real-time gender classication based on facial images. The proposed CNN architecture exhibits a much reduced design complexity when compared with other CNN solutions applied in pattern recognition. The number of processing layers in the CNN is reduced to only four by fusing the convolutional and subsampling layers. Unlike in conventional CNNs, we replace the convolution operation with cross-correlation, hence reducing the computational load. The network is trained using a second-order backpropagation learning algorithm with annealed global learning rates. Performance evaluation of the proposed CNN solution is conducted on two publicly available face databases of SUMS and AT&T. We achieve classication accuracies of 98.75% and 99.38% on the SUMS and AT&T databases, respectively. The neural network is able to process and classify a 32 32 pixel face image in less than 0.27 ms, which corresponds to a very high throughput of over 3700 images per second. Training converges within less than 20 epochs. These results correspond to a superior classication performance, verifying that the proposed CNN is an effective real-time solution for gender recognition. TUBITAK 2016-01-03 Article PeerReviewed Liew, Shan Sung and Mohd. Hani, Mohamed Khalil and Ahmad Radzi, Syafeeza and Bakhteri, Rabia (2016) Gender classication: a convolutional neural network approach. Turkish Journal of Electrical Engineering & Computer Sciences, 24 . pp. 1248-1264. ISSN 1300-0632 http://dx.doi.org/10.3906/elk-1311-58 DOI:10.3906/elk-1311-58 |
institution |
Universiti Teknologi Malaysia |
building |
UTM Library |
collection |
Institutional Repository |
continent |
Asia |
country |
Malaysia |
content_provider |
Universiti Teknologi Malaysia |
content_source |
UTM Institutional Repository |
url_provider |
http://eprints.utm.my/ |
topic |
TK Electrical engineering. Electronics Nuclear engineering |
spellingShingle |
TK Electrical engineering. Electronics Nuclear engineering Liew, Shan Sung Mohd. Hani, Mohamed Khalil Ahmad Radzi, Syafeeza Bakhteri, Rabia Gender classication: a convolutional neural network approach |
description |
An approach using a convolutional neural network (CNN) is proposed for real-time gender classication based on facial images. The proposed CNN architecture exhibits a much reduced design complexity when compared with other CNN solutions applied in pattern recognition. The number of processing layers in the CNN is reduced to only four by fusing the convolutional and subsampling layers. Unlike in conventional CNNs, we replace the convolution operation with cross-correlation, hence reducing the computational load. The network is trained using a second-order backpropagation learning algorithm with annealed global learning rates. Performance evaluation of the proposed CNN solution is conducted on two publicly available face databases of SUMS and AT&T. We achieve classication accuracies of 98.75% and 99.38% on the SUMS and AT&T databases, respectively. The neural network is able to process and classify a 32 32 pixel face image in less than 0.27 ms, which corresponds to a very high throughput of over 3700 images per second. Training converges within less than 20 epochs. These results correspond to a superior classication performance, verifying that the proposed CNN is an effective real-time solution for gender recognition. |
format |
Article |
author |
Liew, Shan Sung Mohd. Hani, Mohamed Khalil Ahmad Radzi, Syafeeza Bakhteri, Rabia |
author_facet |
Liew, Shan Sung Mohd. Hani, Mohamed Khalil Ahmad Radzi, Syafeeza Bakhteri, Rabia |
author_sort |
Liew, Shan Sung |
title |
Gender classication: a convolutional neural network approach |
title_short |
Gender classication: a convolutional neural network approach |
title_full |
Gender classication: a convolutional neural network approach |
title_fullStr |
Gender classication: a convolutional neural network approach |
title_full_unstemmed |
Gender classication: a convolutional neural network approach |
title_sort |
gender classication: a convolutional neural network approach |
publisher |
TUBITAK |
publishDate |
2016 |
url |
http://eprints.utm.my/id/eprint/68390/ http://dx.doi.org/10.3906/elk-1311-58 |
_version_ |
1643655941723258880 |