🧠 Skin Disease Classification Backend

Backend system for skin disease classification using Deep Learning and Machine Learning models. This project integrates image-based prediction (CNN) with metadata analysis (Naive Bayes) to improve diagnostic accuracy.

---

🤗 Live Demo (Hugging Face Space)

Experience real-time skin disease classification directly in your browser.

🚀 Features

• 🖼️ Image classification using CNN (ResNet / EfficientNet)
• 📊 Metadata-based prediction using Naive Bayes
• 🔗 Combined prediction pipeline (Image + Metadata)
• ⚡ FastAPI backend for model serving
• 📁 JSON-based label mapping
• 🔌 Ready for integration with mobile/web applications

---

🏗️ Project Structure

Skin_Classify_Backend/
│
├── main.py            # Entry point (API / inference pipeline)
├── model_def.py       # Model architecture definitions
├── labels.json        # Disease label mapping
├── requirements.txt   # Dependencies
└── .gitignore

---

🧠 Model Pipeline

Input Image
    ↓
CNN Model (ResNet / EfficientNet)
    ↓
Disease Probabilities
    ↓
Combine with Metadata (Naive Bayes)
    ↓
Final Prediction

---

⚙️ Installation

git clone https://github.com/YOUR_USERNAME/Skin_Classify_Backend.git
cd Skin_Classify_Backend

pip install -r requirements.txt

---

▶️ Run the Project

python main.py

---

📊 Supported Diseases

• Eczema
• Psoriasis
• Urticaria
• Folliculitis
• Insect Bite
• Contact Dermatitis

---

🛠️ Tech Stack

• Python
• PyTorch / TensorFlow
• FastAPI
• Scikit-learn
• Pandas / NumPy

---

📌 Status

🚧 Currently under development Some features and models are still being improved.

---

⭐ Future Improvements

• Add LLM for medical explanation (Llama / Meditron)
• Improve model accuracy with more data
• Model comparison (ResNet vs EfficientNet)
• Deploy to cloud (AWS / GCP)
• Add frontend dashboard

---

👨‍💻 Author

• Flupxscop

---

📄 License

This project is for educational and research purposes.