Authors :
Ahmet Haydar Ornek; Mustafa Celik; Murat Ceylan
Volume/Issue :
Volume 6 - 2021, Issue 9 - September
Google Scholar :
http://bitly.ws/gu88
Scribd :
https://bit.ly/2ZD1bys
Abstract :
The image classification has become a wellknown process with the development of deep neural
networks. Although classification studies above 90%
accuracy are realized, their explainable side is still an open
area which means the classification process are not known
by researchers. In this study, we show what a deep neural
network model learns from face images to classify them
into with mask and without mask classes using last
convolutional layers of the model. As a deep neural
network model ResNet-18 was selected and the model was
trained with 18600 balanced face images belonging two
classes and tested with 4540 face images different from
training images. The model's test results are obtained as
95.16% sensitivity, 96.69% specificity, 96.58% accuracy.
With the created activation maps it is clearly seen that the
model learns face structure for images without mask and
mask structure for images with mask.
Keywords :
Classification; Covid-19; Explainable Artificial Intelligence; Transfer Learning
The image classification has become a wellknown process with the development of deep neural
networks. Although classification studies above 90%
accuracy are realized, their explainable side is still an open
area which means the classification process are not known
by researchers. In this study, we show what a deep neural
network model learns from face images to classify them
into with mask and without mask classes using last
convolutional layers of the model. As a deep neural
network model ResNet-18 was selected and the model was
trained with 18600 balanced face images belonging two
classes and tested with 4540 face images different from
training images. The model's test results are obtained as
95.16% sensitivity, 96.69% specificity, 96.58% accuracy.
With the created activation maps it is clearly seen that the
model learns face structure for images without mask and
mask structure for images with mask.
Keywords :
Classification; Covid-19; Explainable Artificial Intelligence; Transfer Learning
