An AI-powered web application that uses a YOLO-based deep learning model to detect and classify urban pollution into six categories: birds, roadkill, trash, graffiti, potholes, and weeds. Users can upload images, and receive real-time predictions to support environmental monitoring and cleanup efforts. By integrating with city surveillance infrastructure — such as cameras and patrol vehicles — the system enables automatic detection and reporting of environmental issues. This proactive approach helps the city respond faster, allocate resources more efficiently, maintain cleaner public spaces, and ultimately reduce long-term maintenance costs while improving community health and urban appeal.

Tech Stack: Python 3.8, OpenCV, PyTorch, Ultralytics, Streamlit

Platforms: Google Colab, VSCode

Collaborators: Harini, Zoya, Aisha

See it work (link in slides): https://docs.google.com/presentation/d/1mZOZZTJv7o5zuEiS23p6px3wd6f8Tld-3JninYe1paU/edit?usp=sharing

Password: 35.201.230.33

Implementation Video: https://youtu.be/EPXhcnL9uAw