Skip to content

N2V Manipulate

Source

N2V manipulation transform for PyTorch.

N2VManipulate

Default augmentation for the N2V model.

This transform expects C(Z)YX dimensions.

Parameters:

  • n2v_manipulate_config (N2VManipulateConfig) –

    N2V manipulation configuration.

  • device (str, default: None ) –

    The device on which operations take place, e.g. "cuda", "cpu" or "mps".

Attributes:

  • masked_pixel_percentage (float) –

    Percentage of pixels to mask.

  • roi_size (int) –

    Size of the replacement area.

  • strategy (Literal[uniform, median]) –

    Replacement strategy, uniform or median.

  • remove_center (bool) –

    Whether to remove central pixel from patch.

  • struct_mask (Optional[StructMaskParameters]) –

    StructN2V mask parameters.

  • rng (Generator) –

    Random number generator.

__call__(batch, *args, **kwargs)

Apply the transform to the image.

Parameters:

  • batch (Tensor) –

    Batch if image patches, 2D or 3D, shape BC(Z)YX.

  • *args (Any, default: () ) –

    Additional arguments, unused.

  • **kwargs (Any, default: {} ) –

    Additional keyword arguments, unused.

Returns:

  • tuple[Tensor, Tensor, Tensor] –

    Masked patch, original patch, and mask.

__init__(n2v_manipulate_config, device=None)

Constructor.

Parameters:

  • n2v_manipulate_config (N2VManipulateConfig) –

    N2V manipulation configuration.

  • device (str, default: None ) –

    The device on which operations take place, e.g. "cuda", "cpu" or "mps".