Comparative Analysis of ResNet, EfficientNet, and MobileNetV2 for Detecting Skin Conditions in Deep Learning Models

Comparing the effectiveness of Skin Condition Detection using ResNet, EfficientNet and MobileNetV2 deep learning models. The 3 skin conditions that have been considered are Acrochordon, Lichenoid Keratosis and Vascular Lesions. The ResNet Model, EfficientNet Model and MobileNetV2 Model have been assessed against one another. ResNet and EfficientNet both predicted the skin condition with an accuracy of 92.1875% which is greater than MobileNetV2, which predicted the skin condition with 90.625% accuracy. Through this research, it has been concluded that ResNet and EfficientNet are significantly better than MobileNetV2 in terms of training models to accurately detect skin conditions.