DEEP LEARNING FRAMEWORK FOR SKIN DISEASE DETECTION

Skin diseases are among the most common health disorders worldwide and require early and accurate diagnosis for effective treatment.Traditional diagnostic methods rely heavily on dermatologists' expertise, making the process subjective, time-consuming, and limited by specialist availability.Recent advancements in deep learning have demonstrated significant potential in automated medical image analysis.This paper proposes an ensemble-based deep learning framework for multi-class skin disease detection using EfficientNetB3 integrated with a Convolutional Block Attention Module (CBAM) and DenseNet121.Transfer learning is applied to leverage pre-trained ImageNet weights for efficient feature extraction.The attention mechanism enhances the model's ability to focus on disease-affected regions, while ensemble learning improves robustness and classification performance.The system is trained and evaluated on a Kaggle skin disease dataset containing eight disease categories, including bacterial, fungal, viral, and parasitic infections.Experimental results show that the proposed ensemble model achieves a classification accuracy of 97.44%, outperforming individual baseline models.The proposed framework can assist dermatologists in early diagnosis and can be integrated into telemedicine applications for remote healthcare support.