convert_original_to_model_coords

lightning_pose.data.utils.convert_original_to_model_coords(batch_dict: MultiviewHeatmapLabeledBatchDict, original_keypoints: Tensor, {'__torchtyping__': True, 'details': ('batch', 'num_views', 'num_keypoints', 2), 'cls_name': 'TensorType'}]) → Tensor, {'__torchtyping__': True, 'details': ('batch', 'num_views', 'num_keypoints', 2,), 'cls_name': 'TensorType'}]

Transform keypoints from original frame coordinates to model input coordinates.