BadmintonAnalyticsCV

Badminton Analytics using CV - so far, highlight creation and automatic score updation using TrackNet and YOLO.

TrackNetV3 Model Repository

YOLOv8 Model Repository

Environment setup:

pip install -r requirements.txt

Steps to run:

Shuttle Tracking Inference using TrackNetV3:

1. Execute the following command line statement

   python3 pre_predict.py --video_file original_short.mp4 (or any other raw footage video) (You can choose to add a --save_dir <dir> argument if you want the predictions to be stored elsewhere and not the default prediction directory)

Using TrackNetV3 predictions for highlights generation and score updation:

1. Execute the following command line statement

   python3 testing.py

After running the above steps, filename_score_clip.mp4 will be created at cwd.