Assistly is your basketball companion app, tracking and analyzing shots in real-time using computer vision. Whether practicing solo or competing with friends, it turns every session into a thrilling experience, complete with:
- Dynamic visual effects
- Customizable audio feedback
- Detailed shooting stats and insights
Uses YOLO for basketballs, rims, and shot events detection.
Captures makes, misses, and shooting percentages.
Exciting animations, such as lightning strikes for streaks.
Choose sound themes:
- Arena Cheers
- Dog Barks
- Minion Reactions
- Cow Moos
Session stats, streaks (hot/cold), and exportable session logs.
To run the project, install these:
- Python 3.6-3.11 (cmu_graphics requirement)
- OpenCV
- PyTorch
- Ultralytics
- NumPy
- CMU Graphics
- Pillow
- Pathlib
git clone https://github.com/matthewxrao/assistly.git
cd assistlypip install -r requirements.txtpython src/main.py-
Start a Session:
- Select automatic/manual modes.
- Customize crowd sounds.
- Begin tracking shots!
-
Analyze in Real-Time:
- View annotated stats and enjoy effects.
- Press 'm' to simulate a made shot, 's' to simulate a missed shot, and 'f' to end your current session and view the session summary.
-
Session Summary:
- Review detailed stats and export data.
Recognized 2nd out of 441 Students at 15-112 Fall 2024 Term Project Showcase.
Youtube Link: https://youtu.be/a7-833gmraQ
- https://www.youtube.com/watch?v=9X3jGGnbcvU & https://automaticaddison.com/extended-kalman-filter-ekf-with-python-code-example/ for help on implementing Extended Kalman Filter
- https://www.101soundboards.com/ for soundboard
- https://vectr.com/ for all graphic assets
- ChatGPT for debugging, particularly with getting pillow to work with my computer