Skin diseases rank among the most prevalent health conditions globally, yet their diagnosis remains challenging, primarily due to the intricate interplay of factors such as skin tone, color variations, and hair presence. The diverse manifestations, initial symptom similarities, and uneven distribution of lesion samples further compound the complexity of accurately classifying these disorders. Deep Convolutional Neural Networks (CNNs) have exhibited remarkable potential in improving the precision of skin disease classification. This paper introduces a novel approach to enhance skin disease classification by fine-tuning a pre-trained Xception model. The fine-tuning process entails the incorporation of supplementary layers into the foundational Xception architecture, with selective weight training. The proposed model builds upon an augmented Xception architecture, thoughtfully incorporating depth wise separable convolutions complemented by Batch Normalization and the versatile RELU/SELU activation functions. Comparative analysis against the original Xception model and earlier architectures unequivocally highlights the substantial enhancement in classification accuracy achieved by our network. Empirical results firmly establish the proposed model's efficiency and reliability relative to previous iterations. Notably, the proposed model surpasses contemporary state-of-the-art models, achieving an exceptional accuracy rate of 99% alongside an F1-score of 97%. This research underscores the potential of fine-tuned Xception-based CNNs in significantly advancing the accuracy and reliability of skin disease classification, offering a robust solution to the multifaceted challenges posed by skin disease diagnosis.

Skin Disease Images Xception Deep learning OpenCV Fine-tune Convolutional neural networks (CNN)