DigiDerma: An Application for Skin Disease Prediction Using Attention-Enabled Deep Transfer Learning Model
摘要
Globally, millions of individuals suffer from skin conditions like melanoma and eczema thereby necessitating early diagnosis. In this work, a deep learning-based system called DigiDerma is developed for efficient skin disease recognition with the help of a lightweight deep learning based on MobileNetV2 aided with squeeze excitation attention module. The Flutter framework is used for application development and PHP for backend API development. The proposed system helps in early diagnosis by enabling quick skin condition assessment through the developed user-friendly mobile application. The proposed system is trained and evaluated on a publicly available dataset, i.e., HAM10000 where it achieves an accuracy of 92% which is found to be higher than that achieved by some recent methods.