Machine Learning experiments on SquashAI data.
This repository contains the following Python Notebooks:
- baseline_court_keypoints (Colab link): estimation of squash court keypoints, trained on the T-box Keypoints dataset
- baseline_court_positions (Colab link): estimation of player positions on the squash court, trained on the Transposed Player Positions dataset
- positions_superposition (Colab link): augment a squash video by superposing court positions of the players, estimated by the model trained in baseline_court_positions