GitHub

Converting basketball video to motion tracking data:  #BigData #DataScience #MachineLearning

  • This code requires using tensorbox to create a model for training.
  • If you’re just interested in running the notebook, a pre-trained model and the test video are located here.
  • Place the video in the data folder and the ckpt file in your output folder

    After creating the model run the steps in the pipeline with your model included on a clip of a game.

  • Currently the model only works in clips that stay on the same half of the court.

player_tracker – Tracking and finding coordinates for players in a youtube video

@KirkDBorne: Converting basketball video to motion tracking data: #BigData #DataScience #MachineLearning

This code requires using tensorbox to create a model for training.

If you’re just interested in running the notebook, a pre-trained model and the test video are located here. Place the video in the data folder and the ckpt file in your output folder

After creating the model run the steps in the pipeline with your model included on a clip of a game. Currently the model only works in clips that stay on the same half of the court.

For more information about the project and future work to be done check out the blog post

You signed in with another tab or window. Reload to refresh your session. You signed out in another tab or window. Reload to refresh your session.

GitHub