Skip to content

Remove performance workaround for mask resize #6729

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
merged 8 commits into from
Oct 10, 2022
Merged

Remove performance workaround for mask resize #6729

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
8 commits
Select commit Hold shift + click to select a range
e650b21
Remove performance workaround for mask resize
datumbox Oct 10, 2022
70d03c2
Fix linter
datumbox Oct 10, 2022
b4fc0f4
bug fixes
datumbox Oct 10, 2022
41c8658
Merge branch 'main' into prototype/remove_resize_workaround
datumbox Oct 10, 2022
f8f1b5e
remove unnecessary import
datumbox Oct 10, 2022
dd3a2b2
Merge branch 'main' into prototype/remove_resize_workaround
datumbox Oct 10, 2022
6a9cbde
Merge branch 'main' into prototype/remove_resize_workaround
datumbox Oct 10, 2022
e8216b0
Fixing linter
datumbox Oct 10, 2022
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
11 changes: 0 additions & 11 deletions test/prototype_transforms_kernel_infos.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -19,7 +19,6 @@
make_video_loaders,
mark_framework_limitation,
TestMark,
VALID_EXTRA_DIMS,
)
from torchvision.prototype import features
from torchvision.transforms.functional_tensor import _max_value as get_max_value
Expand Down Expand Up @@ -215,16 +214,6 @@ def sample_inputs_resize_image_tensor():
):
yield ArgsKwargs(image_loader, size=[min(image_loader.image_size) + 1], interpolation=interpolation)

# We have a speed hack in place for nearest interpolation and single channel images (grayscale)
for image_loader in make_image_loaders(
sizes=["random"],
color_spaces=[features.ColorSpace.GRAY],
extra_dims=VALID_EXTRA_DIMS,
):
yield ArgsKwargs(
image_loader, size=[min(image_loader.image_size) + 1], interpolation=F.InterpolationMode.NEAREST
)

yield ArgsKwargs(make_image_loader(size=(11, 17)), size=20, max_size=25)


Expand Down
39 changes: 7 additions & 32 deletions torchvision/prototype/transforms/functional/_geometry.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -14,12 +14,7 @@
pil_to_tensor,
to_pil_image,
)
from torchvision.transforms.functional_tensor import (
_cast_squeeze_in,
_cast_squeeze_out,
_parse_pad_padding,
interpolate,
)
from torchvision.transforms.functional_tensor import _parse_pad_padding

from ._meta import (
convert_format_bounding_box,
Expand Down Expand Up @@ -130,32 +125,12 @@ def resize_image_tensor(
if image.numel() > 0:
image = image.view(-1, num_channels, old_height, old_width)

# This is a perf hack to avoid slow channels_last upsample code path
# Related issue: https://github.com/pytorch/pytorch/issues/83840
# We are transforming (N, 1, H, W) into (N, 2, H, W) to force to take channels_first path
if image.shape[1] == 1 and interpolation == InterpolationMode.NEAREST:
# Below code is copied from _FT.resize
# This is due to the fact that we need to apply the hack on casted image and not before
# Otherwise, image will be copied while cast to float and interpolate will work on twice more data
image, need_cast, need_squeeze, out_dtype = _cast_squeeze_in(image, [torch.float32, torch.float64])

shape = (image.shape[0], 2, image.shape[2], image.shape[3])
image = image.expand(shape)

image = interpolate(
image, size=[new_height, new_width], mode=interpolation.value, align_corners=None, antialias=False
)

image = image[:, 0, ...]
image = _cast_squeeze_out(image, need_cast=need_cast, need_squeeze=need_squeeze, out_dtype=out_dtype)

else:
image = _FT.resize(
image,
size=[new_height, new_width],
interpolation=interpolation.value,
antialias=antialias,
)
image = _FT.resize(
image,
size=[new_height, new_width],
interpolation=interpolation.value,
antialias=antialias,
)

return image.view(extra_dims + (num_channels, new_height, new_width))

Expand Down