khhamid/plants-diseases-detection

🌿 Plant Diseases Classifier (MobileNetV3 + TFLite)


An optimized plant disease detection web app built using TensorFlow, MobileNetV3, and Streamlit, with quantized TFLite deployment for efficient inference on the web, cloud, and edge devices. This project demonstrates real-world model optimization, containerized deployment, and automated CI/CD to Hugging Face Spaces and Docker Hub.

🚀 Features

• —✅ Fine-tuned MobileNetV3-Small for accurate plant disease recognition
• —⚡ TFLite quantization (INT8) for ultra-fast inference
• —🌐 Streamlit-based web interface for image upload and prediction
• —📦 Dockerized environment for reproducible builds
• —🤗 CI/CD pipeline that automatically pushes to:
• —Docker Hub
• —Hugging Face Models & Spaces

🧠 Model Overview

📂 Project Structure

.
├── models/
│ └── mobilenet_int8.tflite # Quantized model
├── src/
│ └── app.py # Streamlit web app
├── class_names.txt # Class labels
├── requirements.txt # Python dependencies
├── Dockerfile # Container setup
├── .github/workflows/hf_deploy.yml # CI/CD workflow
└── README.md

🧰 Local Setup

Clone the repository

git clone https://github.com/<your-username>/plants-diseases-lite.git
cd plants-diseases-lite

Install dependencies

pip install -r requirements.txt

Run the app

streamlit run src/app.py --server.port=7860

Then open: http://localhost:7860

Docker Deployment

1 Build the image

docker build -t plants-diseases-lite .

2 Run the container

docker run -p 7860:7860 plants-diseases-lite

Open your browser at http://localhost:7860

🤗 Hugging Face Integration

This project includes a full CI/CD pipeline that automatically:

• —Builds the app

• —Pushes the Docker image to Docker Hub

• —Uploads the quantized model to Hugging Face Model Hub

• —Deploys the Streamlit app to Hugging Face Spaces

📈 Results Metric | Keras Model | TFLite INT8 | |-----|------------|-------------| Accuracy | 98% | 98% | Avg Inference Time | - ms | ~2.7 ms | Model Size | 22.4 MB | 1.0 MB |

🌍 Live Demo

Try it live on Hugging Face Spaces 👇 👉 https://huggingface.co/spaces/khhamid/plants-diseases-lite-app

Future Work

• —📱 Deploy on edge & mobile devices using:

• —☁️ Add backend API for large-scale cloud predictions

• —🌾 Expand dataset with real-world agricultural images

• —📷 Enable live camera inference for instant in-field diagnosis

🪴 License

This project is released under the MIT License. You are free to use, modify, and distribute it with attribution.

Refs:

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

⭐ If you like this project, give it a star!

Your feedback helps improve open-source AI for agriculture 🌱