generate_labeled_video

lightning_pose.utils.predictions.generate_labeled_video(video_file: str, preds_df: DataFrame, output_mp4_file: str, confidence_thresh_for_vid: float, colormap: str) → None

Overlay keypoint markers on a video and write the result to disk.

Parameters:

• video_file – path to the source video file.

• preds_df – predictions DataFrame with columns indexed as (scorer, bodypart, coord) where coord is x, y, or likelihood.

• output_mp4_file – path where the labeled video will be saved.

• confidence_thresh_for_vid – keypoints with confidence below this value are not plotted.

• colormap – matplotlib colormap name used to colour each keypoint.