Scientific Programming / 2021 / Article / Tab 1 / Research Article
Real-Time Implementation of AI-Based Face Mask Detection and Social Distancing Measuring System for COVID-19 Prevention Table 1 Comparison of different coronavirus detection methods.
Ref Model used Image type Accuracy (%) Advantages Disadvantages [13 ] CNN-based model Chest X-ray 91.34 A simple but effective CNN model for the detection of COVID-19 A limited number of images available [18 ] DarkCovidNet Chest X-ray 87.02 Able to perform binary and multiclass tasks can be used to diagnose other chest-related diseases including tuberculosis and pneumonia [19 ] COVIDX-Net Chest X-ray 93.3 An open source and accessible to general public [20 ] Self-developed model (CTnet-10) CT 82.1 All the automatic diagnosis presented previously can be used by doctors as an efficient and quite method for coronavirus disease screening VGG-16 89 ResNet-50 60 InceptionV3 53.4 VGG-19 94.52 DenseNet-169 93.15 [21 ] CNN modality CT 99 The proposed modalities can be considered in an efficient diagnosis system for the detection of COVID-19 and other relevant infections The proposed modalities achieve poor performance in some cases especially for the ConvLSTM modality due to the dependency on the previous states in the structure of the ConvLSTM modality Augmented dataset A 99 Augmented dataset B 100 Radiography X-ray 95 Combined dataset (CT scan and X-ray) 99 ConvLSTM CT 99 Augmented dataset A 100 Augmented dataset B 99 Radiography X-ray 88 Combined dataset (CT scan and X-ray) 99