Developed a deep learning system using VGG-16 and ResNet-50 to classify skin diseases from DermNet images, achieving 96.03% and 90.48% validation accuracy respectively. Implemented transfer learning and data augmentation to optimize performance for clinical applications. Demonstrated potential to reduce diagnostic time by 60%, bridging AI and healthcare diagnostics.
Language: Python Libraries: TensorFlow/Keras, VGG-16, ResNet-50, OpenCV, Google Colab Dataset: DermNet (2 distinct skin diseases)