Ftl-CoV19: A Transfer Learning Approach to Detect COVID-19

Table 1

Summary of the above state of the arts.

Author name	References	Highlights and contribution
Mertz	[16]	It focuses on computational-based methodologies and tools to analyze CT scans and chest X-rays like QXR
Pham et al.	[17]	Provides a compilation the state-of-the-art big data application that can aid in COVID-19 outbreak prediction, tracking, diagnosis, and drug discovery
Zheng et al.	[18]	This paper proposed a software-based tool using 3D CT volumes to detect COVID-19 utilizing the pretrained UNet model for lung segmentation
Oh et al.	[19]	An openly accessible deep convolutional neural network platform called COVID-Net with 80% sensitivity
Wang et al.	[20]	DeConVNet required training that consisted of 499 CT scans and taking over 20 hours, plotted ROC and PR curves model obtained a TPR of 0.880
Li and He	[21]	It showcases the advantage of ResNet over the VGG series due to gradient fading in identifying the shortcut connections
Rahimzadeh and Attar	[22]	It provides a classification based on three parameters such as COVID-19, pneumonia, and normal, trained on X-ray images resulting a concatenated neural network of Xception and ResNet50V2
Wang et al.	[23]	InceptionV3-based deep learning model, which results in comparative analysis between the pretrained models such as VGG17, AlexNet16, ResNet19, and NASNet