“…From the studies indicated in Table 7 it is visible that, 61.9% of the studies have achieved more than 95% accuracy. The models used in these studies are ResNet-50 [56] , VGG-16 [56] , [72] , VGG-19 [70] , CapsNet [88] , DensNet121 [75] , MobileNet [82] , MobileNetV2 [83] , InceptionV3 [83] , Xception [83] , InceptionResNetV2 [82] and CNNs developed from scratch [6] , [105] , [107] , [108] , [129] . From this analysis, it can be seen that among the studies that have achieved more than 95% accuracy values, only 38.5% of studies have come up with newly developed CNN architectures.…”