A project that combines Knowledge Graphs Representation, Machine Learning and Computer Vision into the topic of football.
Backend:
- Python and Flask framework
- OpenCV
- TensorFlow Keras
- Apache Jena Fuseki - for hosting the Knowledge Graph
- SparQL: language for interacting with the Knowledge Graph
Frontend:
- HTML
- CSS (and sometimes Bootstrap)
- JavaScript
Create a knowledge graph with all the teams of EURO2024;
- matches to be played, info about venues and about squads will be also known.
Make a presentation page with Venues.
Make presentation pages containing information about National Teams.
Develop a feature that takes as input a picture of two players and can predict
the team which they are part of.
For the last requirement OpenCV was used to allow the user to manually drag the bounding-boxes.
In order to start the program, you need to start Apache Jena Fuseki Server.
Inside Fuseki you need to load the Knowledge Graph.
Then run app.py and launch the given localhost url.
The project was done under the coordination of Prof. Florian Heinrichs of Hochschule Darmstadt of Applied Sciences, Germany.
Also, I couldn't pull this project off without the valuable help of my colleagues:
- Ly Nguyen
- Jan Schmalfuß
- Ben Hamouda