convert_original_to_model_coords

lightning_pose.data.utils.convert_original_to_model_coords(batch_dict: MultiviewHeatmapLabeledBatchDict, original_keypoints: Float[Tensor, 'batch num_views num_keypoints 2']) → Float[Tensor, 'batch num_views num_keypoints 2'][source]: Transform keypoints from original frame coordinates to model input coordinates.