Skip to content
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

Jump to bottom

mlx - image.resize add crop_to_aspect_ratio argument #19699

Merged
merged 1 commit into from
May 13, 2024
Merged

mlx - image.resize add crop_to_aspect_ratio argument #19699

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Jump to
Jump to file
Failed to load files.
Diff view
Diff view
133 changes: 131 additions & 2 deletions keras/src/backend/mlx/image.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -300,20 +300,28 @@ def resize(
size,
interpolation="bilinear",
antialias=False,
crop_to_aspect_ratio=False,
pad_to_aspect_ratio=False,
fill_mode="constant",
fill_value=0.0,
data_format="channels_last",
):
if antialias:
raise NotImplementedError(
"Antialiasing not implemented for the MLX backend"
)

if pad_to_aspect_ratio and crop_to_aspect_ratio:
raise ValueError(
"Only one of `pad_to_aspect_ratio` & `crop_to_aspect_ratio` "
"can be `True`."
)
if interpolation not in AFFINE_TRANSFORM_INTERPOLATIONS.keys():
raise ValueError(
"Invalid value for argument `interpolation`. Expected of one "
f"{set(AFFINE_TRANSFORM_INTERPOLATIONS.keys())}. Received: "
f"interpolation={interpolation}"
)

target_height, target_width = size
size = tuple(size)
image = convert_to_tensor(image)

Expand All @@ -324,6 +332,127 @@ def resize(
f"image.shape={image.shape}"
)

if crop_to_aspect_ratio:
shape = image.shape
if data_format == "channels_last":
height, width = shape[-3], shape[-2]
else:
height, width = shape[-2], shape[-1]
crop_height = int(float(width * target_height) / target_width)
crop_height = min(height, crop_height)
crop_width = int(float(height * target_width) / target_height)
crop_width = min(width, crop_width)
crop_box_hstart = int(float(height - crop_height) / 2)
crop_box_wstart = int(float(width - crop_width) / 2)
if data_format == "channels_last":
if len(image.shape) == 4:
image = image[
:,
crop_box_hstart : crop_box_hstart + crop_height,
crop_box_wstart : crop_box_wstart + crop_width,
:,
]
else:
image = image[
crop_box_hstart : crop_box_hstart + crop_height,
crop_box_wstart : crop_box_wstart + crop_width,
:,
]
else:
if len(image.shape) == 4:
image = image[
:,
:,
crop_box_hstart : crop_box_hstart + crop_height,
crop_box_wstart : crop_box_wstart + crop_width,
]
else:
image = image[
:,
crop_box_hstart : crop_box_hstart + crop_height,
crop_box_wstart : crop_box_wstart + crop_width,
]
elif pad_to_aspect_ratio:
shape = image.shape
batch_size = image.shape[0]
if data_format == "channels_last":
height, width, channels = shape[-3], shape[-2], shape[-1]
else:
channels, height, width = shape[-3], shape[-2], shape[-1]
pad_height = int(float(width * target_height) / target_width)
pad_height = max(height, pad_height)
pad_width = int(float(height * target_width) / target_height)
pad_width = max(width, pad_width)
img_box_hstart = int(float(pad_height - height) / 2)
img_box_wstart = int(float(pad_width - width) / 2)
if data_format == "channels_last":
if len(image.shape) == 4:
padded_img = (
mx.ones(
(
batch_size,
pad_height + height,
pad_width + width,
channels,
),
dtype=image.dtype,
)
* fill_value
)
padded_img[
:,
img_box_hstart : img_box_hstart + height,
img_box_wstart : img_box_wstart + width,
:,
] = image
else:
padded_img = (
mx.ones(
(pad_height + height, pad_width + width, channels),
dtype=image.dtype,
)
* fill_value
)
padded_img[
img_box_hstart : img_box_hstart + height,
img_box_wstart : img_box_wstart + width,
:,
] = image
else:
if len(image.shape) == 4:
padded_img = (
mx.ones(
(
batch_size,
channels,
pad_height + height,
pad_width + width,
),
dtype=image.dtype,
)
* fill_value
)
padded_img[
:,
:,
img_box_hstart : img_box_hstart + height,
img_box_wstart : img_box_wstart + width,
] = image
else:
padded_img = (
mx.ones(
(channels, pad_height + height, pad_width + width),
dtype=image.dtype,
)
* fill_value
)
padded_img[
:,
img_box_hstart : img_box_hstart + height,
img_box_wstart : img_box_wstart + width,
] = image
image = padded_img

# Change to channels_last
if data_format == "channels_first":
image = (
Expand Down
8 changes: 8 additions & 0 deletions keras/src/ops/image_test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -273,6 +273,14 @@ def test_resize(self, interpolation, antialias, data_format):
f"Received: interpolation={interpolation}, "
f"antialias={antialias}."
)
if backend.backend() == "mlx":
if interpolation in ["lanczos3", "lanczos5", "bicubic"]:
self.skipTest(
f"Resizing with interpolation={interpolation} is "
"not supported by the mlx backend. "
)
elif antialias:
self.skipTest("antialias=True not supported by mlx backend.")
# Unbatched case
if data_format == "channels_first":
x = np.random.random((3, 50, 50)) * 255
Expand Down