International Research Journal of Education and Technology

Skin diseases represent a significant global health issue, affecting individuals across diverse age groups and regions. Early and accurate diagnosis is essential for effective management and prevention of complications, yet traditional diagnostic methods relying on dermatologist expertise can be time-consuming, subjective, and less accessible in resource-limited settings. Advances in artificial intelligence, particularly deep learning, have opened new avenues for automating skin disease detection and enhancing diagnostic efficiency. This project explores the application of Convolutional Neural Networks (CNNs) and pre-trained deep learning models, such as VGG16, to classify a range of skin conditions. By training and evaluating various models, this study assesses the viability of deep learning for automated skin disease classification, with attention to accuracy, generalization, and computational feasibility. Results suggest that while CNN architectures can detect relevant patterns, deeper models, like VGG16, offer improved performance in distinguishing complex skin conditions. Further, advanced architectures such as Res-Net and Efficient-Net are anticipated to enhance generalization and efficiency, addressing the limitations found in simpler models. Overall, this project underscores the potential of deep learning models to support dermatological diagnostics and improve access to timely, accurate treatment. The findings lay groundwork for continued research toward developing robust, scalable solutions for automated skin disease detection in clinical and remote settings.

Key Words: VGG16, Convolutional Neural Networks (CNNs), Res-Net and Efficient-Net, Feature extraction.