<![CDATA[Facial skin diseases are a common health issue that significantly affect an individual's quality of life. Early detection through image processing is a crucial step for timely treatment. This study applies Convolutional Neural Network with EfficientNet-B3 architecture to classify five types of facial skin diseases, namely acne, actinic keratosis, basal cell carcinoma, eczema, and rosacea. The model was developed through fine-tuning on an augmented image dataset, with training and testing data splits. Evaluation results show a testing accuracy of 96.61 percent, accompanied by average precision, recall, and F1-score values of 0.97. The confusion matrix indicates high classification performance with minimal errors between classes. This approach proves effective in improving detection accuracy, thus potentially supporting medical personnel in early diagnosis.]]>
Copyrights © 2026
