🩺 SmartSkin AI – Skin Disease Classifier

MobileNetV2 CNN with 82% validation accuracy | TensorFlow + Keras | Streamlit UI | Hugging Face Deployment

Features

• 4-class skin lesion classification (nv, mel, bkl, df)
• Data augmentation + preprocessing
• Offline CPU-only app (Windows ready)
• One-click installer for clients

Live Demo

• Hugging Face Space → SmartSkin Demo
• Streamlit local → streamlit run app/streamlit_app.py

Quick Start

git clone https://github.com/Gourav-512/SmartSkin-AI-Deep-Learning.git
cd SmartSkin-AI-Deep-Learning
python -m venv venv
venv\Scripts\activate
pip install -r requirements.txt
streamlit run app/streamlit_app.py

,,..
-