delta.data.trainGenerator_seg

delta.data.trainGenerator_seg(batch_size: int, img_path: str, mask_path: str, weight_path: Optional[str], target_size: Tuple[int, int] = (256, 32), augment_params: Dict[str, Any] = {}, preload: bool = False, seed: int = 1, crop_windows: bool = False) → Iterator[Tuple[ndarray[Any, dtype[float32]], ndarray[Any, dtype[float32]]]]

Generator for training the segmentation U-Net.

Parameters

batch_sizeint

Batch size, number of training samples to concatenate together.

img_pathstring

Path to folder containing training input images.

mask_pathstring

Path to folder containing training segmentation groundtruth.

weight_pathstring or None.

Path to folder containing weight map images.

target_sizetuple of 2 ints, optional

Input and output image size. The default is (256,32).

augment_paramsdict, optional

Data augmentation parameters. See data_augmentation() doc for more info The default is {}.

preloadbool, optional

Flag to load all training inputs in memory during intialization of the generator. The default is False.

seedint, optional

Seed for numpy’s random generator. see numpy.random.seed() doc The default is 1.

Yields

image_arr4D numpy array of floats

Input images for the U-Net training routine. Dimensions of the tensor are (batch_size, target_size[0], target_size[1], 1)

mask_wei_arr4D numpy array of floats

Output masks and weight maps for the U-Net training routine. Dimensions of the tensor are (batch_size, target_size[0], target_size[1], 2)