This is a web application for image processing. It is based on Flask and Celery. The backend is written in Python and the frontend is written in Next.js.
The main goal of this project is to demonstrate the model poisoning attack and defense. The attack is based on the paper . The defense is based on the paper Defending Against Model Poisoning Attacks.
- Install the required packages
pip install pipenv
pipenv install- Run the virtual environment
pipenv shell- Run celery
cd web/backend
celery -A app.celery worker --loglevel=info- Run the web
python3 ./web/backend/app.py- Run the frontend (http://localhost:3000)
cd web/frontend
npm run devpipenv install <package_name>Todo:
- original image preview
- progress bar (discuss)
- Pipline intergration (attack)
- Pipline intergration (defense)
- Pipline celery get task.info
- backend send image to frontend