TY - GEN
T1 - On the ethnic classification of Pakistani face using deep learning
AU - Jilani, Shelina Khalid
AU - Ugail, Hassan
AU - Bukar, Ali Maina
AU - Logan, Andrew
N1 - Acceptance in SAN
AAM: no embargo
PY - 2019/12/5
Y1 - 2019/12/5
N2 - Demographic-based identification plays an active role in the field of face identification. Over the past decade, machine learning algorithms have been used to investigate challenges surrouding ethnic classification for specific populations, such as African, Asian and Caucasian people. Ethnic classification for individuals of South Asian, Pakistani heritage, however, remains to be addressed. The present paper addresses a two-category (Pakistani Vs Non-Pakistani) classification task from a novel, purpose-built dataset. To the best of our knowledge, this work is the first to report a machine learning ethnic classification task with South Asian (Pakistani) faces. We conduted a series of experiments using deep learning algorithms (ResNet-50, ResNet-101 and ResNet-152) for feature extraction and a linear support vector machine (SVM) for classification. The experimental results demonstrate ResNet-101 achieves the highest performance accuracy of 99.2% for full-face ethnicity classification, followed closely by 91.7% and 95.7% for the nose and mouth respectively.
AB - Demographic-based identification plays an active role in the field of face identification. Over the past decade, machine learning algorithms have been used to investigate challenges surrouding ethnic classification for specific populations, such as African, Asian and Caucasian people. Ethnic classification for individuals of South Asian, Pakistani heritage, however, remains to be addressed. The present paper addresses a two-category (Pakistani Vs Non-Pakistani) classification task from a novel, purpose-built dataset. To the best of our knowledge, this work is the first to report a machine learning ethnic classification task with South Asian (Pakistani) faces. We conduted a series of experiments using deep learning algorithms (ResNet-50, ResNet-101 and ResNet-152) for feature extraction and a linear support vector machine (SVM) for classification. The experimental results demonstrate ResNet-101 achieves the highest performance accuracy of 99.2% for full-face ethnicity classification, followed closely by 91.7% and 95.7% for the nose and mouth respectively.
KW - ethnicity
KW - Pakistani
KW - deep learning
KW - residual network
KW - classification
U2 - 10.1109/CW.2019.00039
DO - 10.1109/CW.2019.00039
M3 - Conference contribution
SN - 9781728122984
SP - 191
EP - 198
BT - 2019 International Conference on Cyberworlds (CW)
PB - IEEE
ER -