🐾 Animal Skin Disease Classifier API

This project is a FastAPI-based backend that classifies common skin conditions in dogs, cats, and cows using pre-trained TensorFlow/Keras models.
It also provides medicine suggestions for the predicted disease using Gemini (via med.py). NOTE: Create and load your own Gemini api key before running the app.

---

🚀 Features

• Image upload support (.jpg, .png, .jpeg)
• Lazy model loading for efficient memory usage
• Supported animals:
  • 🐶 Dogs → Demodicosis, Dermatitis, Fungal Infection, Hypersensitivity, Ringworm, Healthy
  • 🐱 Cats → Flea Allergy, Ringworm, Scabies, Healthy
  • 🐄 Cows → Lumpy Skin, Healthy
• Normalizes inconsistent labels
• Fetches medicine suggestions using Gemini

---

📂 Project Structure

.
├── app.py              # Main FastAPI application
├── model_loader.py     # Handles lazy loading of Keras models
├── utils.py            # Utility functions (e.g., image preprocessing)
├── med.py              # Gemini API integration for medicine suggestions
├── Dog.keras           # Pre-trained dog model
├── cat.keras           # Pre-trained cat model
├── cow.keras           # Pre-trained cow model
├── requirements.txt    # Python dependencies
└── README.md           # Project documentation


---

## ⚙️ Installation & Running

```bash
git clone https://github.com/ankitkr1375/animal_classifier.git
cd animal_classifier

python -m venv venv
# Windows
venv\Scripts\activate
# Mac/Linux
source venv/bin/activate

pip install -r requirements.txt
venv\Scripts\activate
uvicorn main:app --reload
streamlit run app.py