A deep learning model trained on YOLOv8 to track football's movement.

The dataset can be downloaded by running the commands from the commands.txt file in the terminal while in the root directory and by following the instructions. The data can also be downloaded manually from Google Open Image Dataset(https://storage.googleapis.com/openimages/web/index.html)

To annotate the data, we can either download the annotated image or we can even do it manually using open image annotations platforms like CVAT(https://www.cvat.ai/).

Once requirements are met, run the train.py file on fewer numbers on epochs(eg: 1 0r 2) to check for any errors. Once resolved, run the file on more than 50 epochs for good accuracy and deep training.

Using the trained model, run predict.py on the video of choice.