dgs.models.dataset.posetrack21.extract_all_bboxes

dgs.models.dataset.posetrack21.extract_all_bboxes(base_dataset_path: str = './data/PoseTrack21/', anno_dir: str = './posetrack_data/', **kwargs) → None[source]

Given the path to the PoseTrack21 dataset, create a new crops folder containing the image crops and its respective key point coordinates of every bounding box separated by test, train, and validation sets like the images. Within every set is one folder per video ID, in which then lie the image crops. The name of the crops is: {person_id}_{image_id}.jpg.

Parameters:

  • base_dataset_path (FilePath) – The path to the PoseTrack21 dataset directory.

  • anno_dir (FilePath) – The name of the directory containing the folders for the training and test annotations.

Notes

For more kwargs see extract_crops_from_json_annotation() and extract_crops_from_images().