Skip to content

2920 Enhance padding mode for Tensor data #2921

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
Nic-Ma merged 12 commits into from
Sep 12, 2021
Merged

2920 Enhance padding mode for Tensor data #2921

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
12 commits
Select commit Hold shift + click to select a range
42a45e0
Merge pull request #19 from Project-MONAI/master
Nic-Ma Feb 1, 2021
cd16a13
Merge pull request #32 from Project-MONAI/master
Nic-Ma Feb 24, 2021
6f87afd
Merge pull request #180 from Project-MONAI/dev
Nic-Ma Jul 22, 2021
f398298
Merge pull request #214 from Project-MONAI/dev
Nic-Ma Sep 8, 2021
bba466b
Merge pull request #217 from Project-MONAI/dev
Nic-Ma Sep 10, 2021
81b52c8
[DLMED] enhance the pad mode
Nic-Ma Sep 10, 2021
8adcc6e
Merge branch 'dev' into 2920-enhance-tensor-pad
Nic-Ma Sep 11, 2021
e904fed
[DLMED] update all the tensor pad related
Nic-Ma Sep 11, 2021
a06fddc
[DLMED] fix error tests
Nic-Ma Sep 11, 2021
a6bbdb6
[DLMED] fix GPU tests
Nic-Ma Sep 11, 2021
153a7ef
[DLMED] update according to comments
Nic-Ma Sep 12, 2021
14abb46
[DLMED] fix document
Nic-Ma Sep 12, 2021
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 6 additions & 0 deletions docs/source/transforms.rst
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -59,6 +59,12 @@ Vanilla Transforms
Crop and Pad
^^^^^^^^^^^^

`Pad`
"""""
.. autoclass:: Pad
:members:
:special-members: __call__

`SpatialPad`
""""""""""""
.. autoclass:: SpatialPad
Expand Down
4 changes: 3 additions & 1 deletion monai/transforms/__init__.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -18,6 +18,7 @@
CenterSpatialCrop,
CropForeground,
DivisiblePad,
Pad,
RandCropByLabelClasses,
RandCropByPosNegLabel,
RandScaleCrop,
Expand Down Expand Up @@ -48,7 +49,7 @@
DivisiblePadd,
DivisiblePadD,
DivisiblePadDict,
NumpyPadModeSequence,
PadModeSequence,
RandCropByLabelClassesd,
RandCropByLabelClassesD,
RandCropByLabelClassesDict,
Expand Down Expand Up @@ -490,6 +491,7 @@
allow_missing_keys_mode,
compute_divisible_spatial_size,
convert_inverse_interp_mode,
convert_pad_mode,
copypaste_arrays,
create_control_grid,
create_grid,
Expand Down
173 changes: 107 additions & 66 deletions monai/transforms/croppad/array.py
View file
Open in desktop

Large diffs are not rendered by default.

53 changes: 30 additions & 23 deletions monai/transforms/croppad/dictionary.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -49,11 +49,11 @@
weighted_patch_samples,
)
from monai.utils import ImageMetaKey as Key
from monai.utils import Method, NumpyPadMode, ensure_tuple, ensure_tuple_rep, fall_back_tuple
from monai.utils import Method, NumpyPadMode, PytorchPadMode, ensure_tuple, ensure_tuple_rep, fall_back_tuple
from monai.utils.enums import InverseKeys

__all__ = [
"NumpyPadModeSequence",
"PadModeSequence",
"SpatialPadd",
"BorderPadd",
"DivisiblePadd",
Expand Down Expand Up @@ -99,6 +99,7 @@
]

NumpyPadModeSequence = Union[Sequence[Union[NumpyPadMode, str]], NumpyPadMode, str]
PadModeSequence = Union[Sequence[Union[NumpyPadMode, PytorchPadMode, str]], NumpyPadMode, PytorchPadMode, str]


class SpatialPadd(MapTransform, InvertibleTransform):
Expand All @@ -114,9 +115,9 @@ def __init__(
keys: KeysCollection,
spatial_size: Union[Sequence[int], int],
method: Union[Method, str] = Method.SYMMETRIC,
mode: NumpyPadModeSequence = NumpyPadMode.CONSTANT,
mode: PadModeSequence = NumpyPadMode.CONSTANT,
allow_missing_keys: bool = False,
**np_kwargs,
**kwargs,
) -> None:
"""
Args:
Expand All @@ -129,19 +130,21 @@ def __init__(
the spatial size of output data will be [32, 30, 30].
method: {``"symmetric"``, ``"end"``}
Pad image symmetrically on every side or only pad at the end sides. Defaults to ``"symmetric"``.
mode: {``"constant"``, ``"edge"``, ``"linear_ramp"``, ``"maximum"``, ``"mean"``,
``"median"``, ``"minimum"``, ``"reflect"``, ``"symmetric"``, ``"wrap"``, ``"empty"``}
mode: available modes for numpy array:{``"constant"``, ``"edge"``, ``"linear_ramp"``, ``"maximum"``,
``"mean"``, ``"median"``, ``"minimum"``, ``"reflect"``, ``"symmetric"``, ``"wrap"``, ``"empty"``}
available modes for PyTorch Tensor: {``"constant"``, ``"reflect"``, ``"replicate"``, ``"circular"``}.
One of the listed string values or a user supplied function. Defaults to ``"constant"``.
See also: https://numpy.org/doc/1.18/reference/generated/numpy.pad.html
https://pytorch.org/docs/stable/generated/torch.nn.functional.pad.html
It also can be a sequence of string, each element corresponds to a key in ``keys``.
allow_missing_keys: don't raise exception if key is missing.
np_kwargs: other args for `np.pad` API, note that `np.pad` treats channel dimension as the first dimension.
more details: https://numpy.org/doc/1.18/reference/generated/numpy.pad.html
kwargs: other arguments for the `np.pad` or `torch.pad` function.
note that `np.pad` treats channel dimension as the first dimension.

"""
super().__init__(keys, allow_missing_keys)
self.mode = ensure_tuple_rep(mode, len(self.keys))
self.padder = SpatialPad(spatial_size, method, **np_kwargs)
self.padder = SpatialPad(spatial_size, method, **kwargs)

def __call__(self, data: Mapping[Hashable, NdarrayTensor]) -> Dict[Hashable, NdarrayTensor]:
d = dict(data)
Expand Down Expand Up @@ -183,9 +186,9 @@ def __init__(
self,
keys: KeysCollection,
spatial_border: Union[Sequence[int], int],
mode: NumpyPadModeSequence = NumpyPadMode.CONSTANT,
mode: PadModeSequence = NumpyPadMode.CONSTANT,
allow_missing_keys: bool = False,
**np_kwargs,
**kwargs,
) -> None:
"""
Args:
Expand All @@ -202,19 +205,21 @@ def __init__(
pad bottom of H dim with 2, pad left of W dim with 3, pad right of W dim with 4.
the result shape is [1, 7, 11].

mode: {``"constant"``, ``"edge"``, ``"linear_ramp"``, ``"maximum"``, ``"mean"``,
``"median"``, ``"minimum"``, ``"reflect"``, ``"symmetric"``, ``"wrap"``, ``"empty"``}
mode: available modes for numpy array:{``"constant"``, ``"edge"``, ``"linear_ramp"``, ``"maximum"``,
``"mean"``, ``"median"``, ``"minimum"``, ``"reflect"``, ``"symmetric"``, ``"wrap"``, ``"empty"``}
available modes for PyTorch Tensor: {``"constant"``, ``"reflect"``, ``"replicate"``, ``"circular"``}.
One of the listed string values or a user supplied function. Defaults to ``"constant"``.
See also: https://numpy.org/doc/1.18/reference/generated/numpy.pad.html
https://pytorch.org/docs/stable/generated/torch.nn.functional.pad.html
It also can be a sequence of string, each element corresponds to a key in ``keys``.
allow_missing_keys: don't raise exception if key is missing.
np_kwargs: other args for `np.pad` API, note that `np.pad` treats channel dimension as the first dimension.
more details: https://numpy.org/doc/1.18/reference/generated/numpy.pad.html
kwargs: other arguments for the `np.pad` or `torch.pad` function.
note that `np.pad` treats channel dimension as the first dimension.

"""
super().__init__(keys, allow_missing_keys)
self.mode = ensure_tuple_rep(mode, len(self.keys))
self.padder = BorderPad(spatial_border=spatial_border, **np_kwargs)
self.padder = BorderPad(spatial_border=spatial_border, **kwargs)

def __call__(self, data: Mapping[Hashable, NdarrayTensor]) -> Dict[Hashable, NdarrayTensor]:
d = dict(data)
Expand Down Expand Up @@ -260,10 +265,10 @@ def __init__(
self,
keys: KeysCollection,
k: Union[Sequence[int], int],
mode: NumpyPadModeSequence = NumpyPadMode.CONSTANT,
mode: PadModeSequence = NumpyPadMode.CONSTANT,
method: Union[Method, str] = Method.SYMMETRIC,
allow_missing_keys: bool = False,
**np_kwargs,
**kwargs,
) -> None:
"""
Args:
Expand All @@ -272,23 +277,25 @@ def __init__(
k: the target k for each spatial dimension.
if `k` is negative or 0, the original size is preserved.
if `k` is an int, the same `k` be applied to all the input spatial dimensions.
mode: {``"constant"``, ``"edge"``, ``"linear_ramp"``, ``"maximum"``, ``"mean"``,
``"median"``, ``"minimum"``, ``"reflect"``, ``"symmetric"``, ``"wrap"``, ``"empty"``}
mode: available modes for numpy array:{``"constant"``, ``"edge"``, ``"linear_ramp"``, ``"maximum"``,
``"mean"``, ``"median"``, ``"minimum"``, ``"reflect"``, ``"symmetric"``, ``"wrap"``, ``"empty"``}
available modes for PyTorch Tensor: {``"constant"``, ``"reflect"``, ``"replicate"``, ``"circular"``}.
One of the listed string values or a user supplied function. Defaults to ``"constant"``.
See also: https://numpy.org/doc/1.18/reference/generated/numpy.pad.html
https://pytorch.org/docs/stable/generated/torch.nn.functional.pad.html
It also can be a sequence of string, each element corresponds to a key in ``keys``.
method: {``"symmetric"``, ``"end"``}
Pad image symmetrically on every side or only pad at the end sides. Defaults to ``"symmetric"``.
allow_missing_keys: don't raise exception if key is missing.
np_kwargs: other args for `np.pad` API, note that `np.pad` treats channel dimension as the first dimension.
more details: https://numpy.org/doc/1.18/reference/generated/numpy.pad.html
kwargs: other arguments for the `np.pad` or `torch.pad` function.
note that `np.pad` treats channel dimension as the first dimension.

See also :py:class:`monai.transforms.SpatialPad`

"""
super().__init__(keys, allow_missing_keys)
self.mode = ensure_tuple_rep(mode, len(self.keys))
self.padder = DivisiblePad(k=k, method=method, **np_kwargs)
self.padder = DivisiblePad(k=k, method=method, **kwargs)

def __call__(self, data: Mapping[Hashable, NdarrayTensor]) -> Dict[Hashable, NdarrayTensor]:
d = dict(data)
Expand Down
Loading