dnrops
/
transformers
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Move center_crop to BaseImageProcessor (#25122)

This commit is contained in:
amyeroberts 2023-07-26 18:30:38 +01:00 committed by GitHub
parent 659829b6ae
commit 1689aea733
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
18 changed files with 29 additions and 419 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

26
src/transformers/image_processing_utils.py
View File

@ -23,7 +23,7 @@ import numpy as np
from .dynamic_module_utils import custom_object_save
from .feature_extraction_utils import BatchFeature as BaseBatchFeature
from .image_transforms import normalize, rescale
from .image_transforms import center_crop, normalize, rescale
from .image_utils import ChannelDimension
from .utils import (
IMAGE_PROCESSOR_NAME,
@ -571,6 +571,30 @@ class BaseImageProcessor(ImageProcessingMixin):
"""
return normalize(image, mean=mean, std=std, data_format=data_format, **kwargs)
def center_crop(
self,
image: np.ndarray,
size: Dict[str, int],
data_format: Optional[Union[str, ChannelDimension]] = None,
**kwargs,
) -> np.ndarray:
"""
Center crop an image to `(size["height"], size["width"])`. If the input size is smaller than `crop_size` along
any edge, the image is padded with 0's and then center cropped.
Args:
image (`np.ndarray`):
Image to center crop.
size (`Dict[str, int]`):
Size of the output image.
data_format (`str` or `ChannelDimension`, *optional*):
The channel dimension format of the image. If not provided, it will be the same as the input image.
"""
size = get_size_dict(size)
if "height" not in size or "width" not in size:
raise ValueError(f"The size dictionary must have keys 'height' and 'width'. Got {size.keys()}")
return center_crop(image, size=(size["height"], size["width"]), data_format=data_format, **kwargs)
VALID_SIZE_DICT_KEYS = ({"height", "width"}, {"shortest_edge"}, {"shortest_edge", "longest_edge"}, {"longest_edge"})

24
src/transformers/models/beit/image_processing_beit.py
View File

@ -20,7 +20,7 @@ from typing import Any, Dict, List, Optional, Tuple, Union
import numpy as np
from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
from ...image_transforms import center_crop, resize, to_channel_dimension_format
from ...image_transforms import resize, to_channel_dimension_format
from ...image_utils import (
IMAGENET_STANDARD_MEAN,
IMAGENET_STANDARD_STD,
@ -167,28 +167,6 @@ class BeitImageProcessor(BaseImageProcessor):
image, size=(size["height"], size["width"]), resample=resample, data_format=data_format, **kwargs
)
def center_crop(
self,
image: np.ndarray,
size: Dict[str, int],
data_format: Optional[Union[str, ChannelDimension]] = None,
**kwargs,
) -> np.ndarray:
"""
Center crop an image to (size["height"], size["width"]). If the input size is smaller than `size` along any
edge, the image is padded with 0's and then center cropped.
Args:
image (`np.ndarray`):
Image to center crop.
size (`Dict[str, int]`):
Size of the output image.
data_format (`str` or `ChannelDimension`, *optional*):
The channel dimension format of the image. If not provided, it will be the same as the input image.
"""
size = get_size_dict(size, default_to_square=True, param_name="size")
return center_crop(image, size=(size["height"], size["width"]), data_format=data_format, **kwargs)
def reduce_label(self, label: ImageInput) -> np.ndarray:
label = to_numpy_array(label)
# Avoid using underflow conversion

25
src/transformers/models/bit/image_processing_bit.py
View File

@ -20,7 +20,6 @@ import numpy as np
from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
from ...image_transforms import (
center_crop,
convert_to_rgb,
get_resize_output_image_size,
resize,
@ -147,30 +146,6 @@ class BitImageProcessor(BaseImageProcessor):
output_size = get_resize_output_image_size(image, size=size["shortest_edge"], default_to_square=False)
return resize(image, size=output_size, resample=resample, data_format=data_format, **kwargs)
def center_crop(
self,
image: np.ndarray,
size: Dict[str, int],
data_format: Optional[Union[str, ChannelDimension]] = None,
**kwargs,
) -> np.ndarray:
"""
Center crop an image. If the image is too small to be cropped to the size given, it will be padded (so the
returned result will always be of size `size`).
Args:
image (`np.ndarray`):
Image to center crop.
size (`Dict[str, int]`):
Size of the output image in the form of a dictionary with keys `height` and `width`.
data_format (`str` or `ChannelDimension`, *optional*):
The channel dimension format of the image. If not provided, it will be the same as the input image.
"""
size = get_size_dict(size)
if "height" not in size or "width" not in size:
raise ValueError(f"The `size` parameter must contain the keys (height, width). Got {size.keys()}")
return center_crop(image, size=(size["height"], size["width"]), data_format=data_format, **kwargs)
def preprocess(
self,
images: ImageInput,

23
src/transformers/models/chinese_clip/image_processing_chinese_clip.py
View File

@ -20,7 +20,6 @@ import numpy as np
from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
from ...image_transforms import (
center_crop,
convert_to_rgb,
get_resize_output_image_size,
resize,
@ -147,28 +146,6 @@ class ChineseCLIPImageProcessor(BaseImageProcessor):
)
return resize(image, size=output_size, resample=resample, data_format=data_format, **kwargs)
def center_crop(
self,
image: np.ndarray,
size: Dict[str, int],
data_format: Optional[Union[str, ChannelDimension]] = None,
**kwargs,
) -> np.ndarray:
"""
Center crop an image. If the image is too small to be cropped to the size given, it will be padded (so the
returned result will always be of size `size`).
Args:
image (`np.ndarray`):
Image to center crop.
size (`Dict[str, int]`):
Size of the output image in the form of a dictionary with keys `height` and `width`.
data_format (`str` or `ChannelDimension`, *optional*):
The channel dimension format of the image. If not provided, it will be the same as the input image.
"""
size = get_size_dict(size)
return center_crop(image, size=(size["height"], size["width"]), data_format=data_format, **kwargs)
def preprocess(
self,
images: ImageInput,

25
src/transformers/models/clip/image_processing_clip.py
View File

@ -20,7 +20,6 @@ import numpy as np
from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
from ...image_transforms import (
center_crop,
convert_to_rgb,
get_resize_output_image_size,
resize,
@ -147,30 +146,6 @@ class CLIPImageProcessor(BaseImageProcessor):
output_size = get_resize_output_image_size(image, size=size["shortest_edge"], default_to_square=False)
return resize(image, size=output_size, resample=resample, data_format=data_format, **kwargs)
def center_crop(
self,
image: np.ndarray,
size: Dict[str, int],
data_format: Optional[Union[str, ChannelDimension]] = None,
**kwargs,
) -> np.ndarray:
"""
Center crop an image. If the image is too small to be cropped to the size given, it will be padded (so the
returned result will always be of size `size`).
Args:
image (`np.ndarray`):
Image to center crop.
size (`Dict[str, int]`):
Size of the output image in the form of a dictionary with keys `height` and `width`.
data_format (`str` or `ChannelDimension`, *optional*):
The channel dimension format of the image. If not provided, it will be the same as the input image.
"""
size = get_size_dict(size)
if "height" not in size or "width" not in size:
raise ValueError(f"The `size` parameter must contain the keys (height, width). Got {size.keys()}")
return center_crop(image, size=(size["height"], size["width"]), data_format=data_format, **kwargs)
def preprocess(
self,
images: ImageInput,

26
src/transformers/models/deit/image_processing_deit.py
View File

@ -19,7 +19,7 @@ from typing import Dict, List, Optional, Union
import numpy as np
from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
from ...image_transforms import center_crop, resize, to_channel_dimension_format
from ...image_transforms import resize, to_channel_dimension_format
from ...image_utils import (
IMAGENET_STANDARD_MEAN,
IMAGENET_STANDARD_STD,
@ -135,30 +135,6 @@ class DeiTImageProcessor(BaseImageProcessor):
image, size=(size["height"], size["width"]), resample=resample, data_format=data_format, **kwargs
)
def center_crop(
self,
image: np.ndarray,
size: Dict[str, int],
data_format: Optional[Union[str, ChannelDimension]] = None,
**kwargs,
) -> np.ndarray:
"""
Center crop an image to `(crop_size["height"], crop_size["width"])`. If the input size is smaller than
`crop_size` along any edge, the image is padded with 0's and then center cropped.
Args:
image (`np.ndarray`):
Image to center crop.
size (`Dict[str, int]`):
Size of the output image.
data_format (`str` or `ChannelDimension`, *optional*):
The channel dimension format of the image. If not provided, it will be the same as the input image.
"""
size = get_size_dict(size)
if "height" not in size or "width" not in size:
raise ValueError(f"The size dictionary must have keys 'height' and 'width'. Got {size.keys()}")
return center_crop(image, size=(size["height"], size["width"]), data_format=data_format, **kwargs)
def preprocess(
self,
images: ImageInput,

25
src/transformers/models/efficientformer/image_processing_efficientformer.py
View File

@ -20,7 +20,6 @@ import numpy as np
from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
from ...image_transforms import (
center_crop,
get_resize_output_image_size,
resize,
to_channel_dimension_format,
@ -149,30 +148,6 @@ class EfficientFormerImageProcessor(BaseImageProcessor):
raise ValueError(f"Size must contain 'height' and 'width' keys or 'shortest_edge' key. Got {size.keys()}")
return resize(image, size=size, resample=resample, data_format=data_format, **kwargs)
def center_crop(
self,
image: np.ndarray,
size: Dict[str, int],
data_format: Optional[Union[str, ChannelDimension]] = None,
**kwargs,
) -> np.ndarray:
"""
Center crop an image. If the image is too small to be cropped to the size given, it will be padded (so the
returned result will always be of size `size`).
Args:
image (`np.ndarray`):
Image to center crop.
size (`Dict[str, int]`):
Size of the output image in the form of a dictionary with keys `height` and `width`.
data_format (`str` or `ChannelDimension`, *optional*):
The channel dimension format of the image. If not provided, it will be the same as the input image.
"""
size = get_size_dict(size)
if "height" not in size or "width" not in size:
raise ValueError(f"The `size` parameter must contain the keys (height, width). Got {size.keys()}")
return center_crop(image, size=(size["height"], size["width"]), data_format=data_format, **kwargs)
def preprocess(
self,
images: ImageInput,

26
src/transformers/models/efficientnet/image_processing_efficientnet.py
View File

@ -19,7 +19,7 @@ from typing import Dict, List, Optional, Union
import numpy as np
from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
from ...image_transforms import center_crop, rescale, resize, to_channel_dimension_format
from ...image_transforms import rescale, resize, to_channel_dimension_format
from ...image_utils import (
IMAGENET_STANDARD_MEAN,
IMAGENET_STANDARD_STD,
@ -144,30 +144,6 @@ class EfficientNetImageProcessor(BaseImageProcessor):
image, size=(size["height"], size["width"]), resample=resample, data_format=data_format, **kwargs
)
def center_crop(
self,
image: np.ndarray,
size: Dict[str, int],
data_format: Optional[Union[str, ChannelDimension]] = None,
**kwargs,
) -> np.ndarray:
"""
Center crop an image to `(crop_size["height"], crop_size["width"])`. If the input size is smaller than
`crop_size` along any edge, the image is padded with 0's and then center cropped.
Args:
image (`np.ndarray`):
Image to center crop.
size (`Dict[str, int]`):
Size of the output image.
data_format (`str` or `ChannelDimension`, *optional*):
The channel dimension format of the image. If not provided, it will be the same as the input image.
"""
size = get_size_dict(size)
if "height" not in size or "width" not in size:
raise ValueError(f"The size dictionary must have keys 'height' and 'width'. Got {size.keys()}")
return center_crop(image, size=(size["height"], size["width"]), data_format=data_format, **kwargs)
def rescale(
self,
image: np.ndarray,

26
src/transformers/models/flava/image_processing_flava.py
View File

@ -22,7 +22,7 @@ from typing import Any, Dict, Iterable, List, Optional, Tuple, Union
import numpy as np
from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
from ...image_transforms import center_crop, resize, to_channel_dimension_format
from ...image_transforms import resize, to_channel_dimension_format
from ...image_utils import (
OPENAI_CLIP_MEAN,
OPENAI_CLIP_STD,
@ -359,30 +359,6 @@ class FlavaImageProcessor(BaseImageProcessor):
image, size=(size["height"], size["width"]), resample=resample, data_format=data_format, **kwargs
)
def center_crop(
self,
image: np.ndarray,
size: Dict[str, int],
data_format: Optional[Union[str, ChannelDimension]] = None,
**kwargs,
) -> np.ndarray:
"""
Center crop an image to `(size["height"], size["width"])`. If the input size is smaller than `crop_size` along
any edge, the image is padded with 0's and then center cropped.
Args:
image (`np.ndarray`):
Image to center crop.
size (`Dict[str, int]`):
Size of the output image.
data_format (`str` or `ChannelDimension`, *optional*):
The channel dimension format of the image. If not provided, it will be the same as the input image.
"""
size = get_size_dict(size)
if "height" not in size or "width" not in size:
raise ValueError(f"The size dictionary must contain 'height' and 'width' keys. Got {size.keys()}")
return center_crop(image, size=(size["height"], size["width"]), data_format=data_format, **kwargs)
def map_pixels(self, image: np.ndarray) -> np.ndarray:
return (1 - 2 * LOGIT_LAPLACE_EPS) * image + LOGIT_LAPLACE_EPS

24
src/transformers/models/levit/image_processing_levit.py
View File

@ -20,7 +20,6 @@ import numpy as np
from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
from ...image_transforms import (
center_crop,
get_resize_output_image_size,
resize,
to_channel_dimension_format,
@ -159,29 +158,6 @@ class LevitImageProcessor(BaseImageProcessor):
image, size=(size_dict["height"], size_dict["width"]), resample=resample, data_format=data_format, **kwargs
)
def center_crop(
self,
image: np.ndarray,
size: Dict[str, int],
data_format: Optional[Union[str, ChannelDimension]] = None,
**kwargs,
) -> np.ndarray:
"""
Center crop an image.
Args:
image (`np.ndarray`):
Image to center crop.
size (`Dict[str, int]`):
Dict `{"height": int, "width": int}` specifying the size of the output image after cropping.
data_format (`str` or `ChannelDimension`, *optional*):
The channel dimension format of the image. If not provided, it will be the same as the input image.
"""
size = get_size_dict(size)
if "height" not in size or "width" not in size:
raise ValueError(f"Size dict must have keys 'height' and 'width'. Got {size.keys()}")
return center_crop(image, size=(size["height"], size["width"]), data_format=data_format, **kwargs)
def preprocess(
self,
images: ImageInput,

23
src/transformers/models/mobilenet_v1/image_processing_mobilenet_v1.py
View File

@ -20,7 +20,6 @@ import numpy as np
from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
from ...image_transforms import (
center_crop,
get_resize_output_image_size,
resize,
to_channel_dimension_format,
@ -140,28 +139,6 @@ class MobileNetV1ImageProcessor(BaseImageProcessor):
output_size = get_resize_output_image_size(image, size=size["shortest_edge"], default_to_square=False)
return resize(image, size=output_size, resample=resample, data_format=data_format, **kwargs)
def center_crop(
self,
image: np.ndarray,
size: Dict[str, int],
data_format: Optional[Union[str, ChannelDimension]] = None,
**kwargs,
) -> np.ndarray:
"""
Center crop an image to (size["height"], size["width"]). If the input size is smaller than `size` along any
edge, the image is padded with 0's and then center cropped.
Args:
image (`np.ndarray`):
Image to center crop.
size (`Dict[str, int]`):
Size of the output image.
data_format (`str` or `ChannelDimension`, *optional*):
The channel dimension format of the image. If not provided, it will be the same as the input image.
"""
size = get_size_dict(size)
return center_crop(image, size=(size["height"], size["width"]), data_format=data_format, **kwargs)
def preprocess(
self,
images: ImageInput,

25
src/transformers/models/mobilenet_v2/image_processing_mobilenet_v2.py
View File

@ -20,7 +20,6 @@ import numpy as np
from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
from ...image_transforms import (
center_crop,
get_resize_output_image_size,
resize,
to_channel_dimension_format,
@ -144,30 +143,6 @@ class MobileNetV2ImageProcessor(BaseImageProcessor):
output_size = get_resize_output_image_size(image, size=size["shortest_edge"], default_to_square=False)
return resize(image, size=output_size, resample=resample, data_format=data_format, **kwargs)
def center_crop(
self,
image: np.ndarray,
size: Dict[str, int],
data_format: Optional[Union[str, ChannelDimension]] = None,
**kwargs,
) -> np.ndarray:
"""
Center crop an image to (size["height"], size["width"]). If the input size is smaller than `size` along any
edge, the image is padded with 0's and then center cropped.
Args:
image (`np.ndarray`):
Image to center crop.
size (`Dict[str, int]`):
Size of the output image.
data_format (`str` or `ChannelDimension`, *optional*):
The channel dimension format of the image. If not provided, it will be the same as the input image.
"""
size = get_size_dict(size)
if "height" not in size or "width" not in size:
raise ValueError(f"The `size` parameter must contain the keys `height` and `width`. Got {size.keys()}")
return center_crop(image, size=(size["height"], size["width"]), data_format=data_format, **kwargs)
def preprocess(
self,
images: ImageInput,

25
src/transformers/models/mobilevit/image_processing_mobilevit.py
View File

@ -20,7 +20,6 @@ import numpy as np
from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
from ...image_transforms import (
center_crop,
flip_channel_order,
get_resize_output_image_size,
resize,
@ -136,30 +135,6 @@ class MobileViTImageProcessor(BaseImageProcessor):
output_size = get_resize_output_image_size(image, size=size["shortest_edge"], default_to_square=False)
return resize(image, size=output_size, resample=resample, data_format=data_format, **kwargs)
def center_crop(
self,
image: np.ndarray,
size: Dict[str, int],
data_format: Optional[Union[str, ChannelDimension]] = None,
**kwargs,
) -> np.ndarray:
"""
Center crop an image to size `(size["height], size["width"])`. If the input size is smaller than `size` along
any edge, the image is padded with 0's and then center cropped.
Args:
image (`np.ndarray`):
Image to center crop.
size (`Dict[str, int]`):
Size of the output image.
data_format (`str` or `ChannelDimension`, *optional*):
The channel dimension format of the image. If not provided, it will be the same as the input image.
"""
size = get_size_dict(size)
if "height" not in size or "width" not in size:
raise ValueError(f"The `size` dictionary must contain the keys `height` and `width`. Got {size.keys()}")
return center_crop(image, size=(size["height"], size["width"]), data_format=data_format, **kwargs)
def flip_channel_order(
self, image: np.ndarray, data_format: Optional[Union[str, ChannelDimension]] = None
) -> np.ndarray:

25
src/transformers/models/poolformer/image_processing_poolformer.py
View File

@ -20,7 +20,6 @@ import numpy as np
from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
from ...image_transforms import (
center_crop,
get_resize_output_image_size,
resize,
to_channel_dimension_format,
@ -193,30 +192,6 @@ class PoolFormerImageProcessor(BaseImageProcessor):
return resize(image, size=output_size, resample=resample, data_format=data_format, **kwargs)
def center_crop(
self,
image: np.ndarray,
size: Dict[str, int],
data_format: Optional[Union[str, ChannelDimension]] = None,
**kwargs,
) -> np.ndarray:
"""
Center crop an image to (size["height"], size["width"]). If the input size is smaller than `crop_size` along
any edge, the image is padded with 0's and then center cropped.
Args:
image (`np.ndarray`):
Image to center crop.
size (`Dict[str, int]`):
Size of the output image.
data_format (`str` or `ChannelDimension`, *optional*):
The channel dimension format of the image. If not provided, it will be the same as the input image.
"""
size = get_size_dict(size)
if "height" not in size or "width" not in size:
raise ValueError(f"size must contain 'height' and 'width' as keys. Got {size.keys()}")
return center_crop(image, size=(size["height"], size["width"]), data_format=data_format, **kwargs)
def preprocess(
self,
images: ImageInput,

25
src/transformers/models/tvlt/image_processing_tvlt.py
View File

@ -19,7 +19,6 @@ import numpy as np
from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
from ...image_transforms import (
center_crop,
get_resize_output_image_size,
resize,
to_channel_dimension_format,
@ -182,30 +181,6 @@ class TvltImageProcessor(BaseImageProcessor):
raise ValueError(f"Size must have 'height' and 'width' or 'shortest_edge' as keys. Got {size.keys()}")
return resize(image, size=output_size, resample=resample, data_format=data_format, **kwargs)
def center_crop(
self,
image: np.ndarray,
size: Dict[str, int],
data_format: Optional[Union[str, ChannelDimension]] = None,
**kwargs,
) -> np.ndarray:
"""
Center crop an image to `(size["height"], size["width"])`. If the input size is smaller than `size` along any
edge, the image is padded with 0's and then center cropped.
Args:
image (`np.ndarray`):
Image to center crop.
size (`Dict[str, int]`):
Size of the output image.
data_format (`str` or `ChannelDimension`, *optional*):
The channel dimension format of the image. If not provided, it will be the same as the input image.
"""
size = get_size_dict(size)
if "height" not in size or "width" not in size:
raise ValueError(f"Size must have 'height' and 'width' as keys. Got {size.keys()}")
return center_crop(image, size=(size["height"], size["width"]), data_format=data_format, **kwargs)
def _preprocess_image(
self,
image: ImageInput,

25
src/transformers/models/videomae/image_processing_videomae.py
View File

@ -20,7 +20,6 @@ import numpy as np
from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
from ...image_transforms import (
center_crop,
get_resize_output_image_size,
resize,
to_channel_dimension_format,
@ -161,30 +160,6 @@ class VideoMAEImageProcessor(BaseImageProcessor):
raise ValueError(f"Size must have 'height' and 'width' or 'shortest_edge' as keys. Got {size.keys()}")
return resize(image, size=output_size, resample=resample, data_format=data_format, **kwargs)
def center_crop(
self,
image: np.ndarray,
size: Dict[str, int],
data_format: Optional[Union[str, ChannelDimension]] = None,
**kwargs,
) -> np.ndarray:
"""
Center crop an image to `(size["height"], size["width"])`. If the input size is smaller than `size` along any
edge, the image is padded with 0's and then center cropped.
Args:
image (`np.ndarray`):
Image to center crop.
size (`Dict[str, int]`):
Size of the output image.
data_format (`str` or `ChannelDimension`, *optional*):
The channel dimension format of the image. If not provided, it will be the same as the input image.
"""
size = get_size_dict(size)
if "height" not in size or "width" not in size:
raise ValueError(f"Size must have 'height' and 'width' as keys. Got {size.keys()}")
return center_crop(image, size=(size["height"], size["width"]), data_format=data_format, **kwargs)
def _preprocess_image(
self,
image: ImageInput,

25
src/transformers/models/vit_hybrid/image_processing_vit_hybrid.py
View File

@ -20,7 +20,6 @@ import numpy as np
from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
from ...image_transforms import (
center_crop,
convert_to_rgb,
get_resize_output_image_size,
resize,
@ -147,30 +146,6 @@ class ViTHybridImageProcessor(BaseImageProcessor):
output_size = get_resize_output_image_size(image, size=size["shortest_edge"], default_to_square=False)
return resize(image, size=output_size, resample=resample, data_format=data_format, **kwargs)
def center_crop(
self,
image: np.ndarray,
size: Dict[str, int],
data_format: Optional[Union[str, ChannelDimension]] = None,
**kwargs,
) -> np.ndarray:
"""
Center crop an image. If the image is too small to be cropped to the size given, it will be padded (so the
returned result will always be of size `size`).
Args:
image (`np.ndarray`):
Image to center crop.
size (`Dict[str, int]`):
Size of the output image in the form of a dictionary with keys `height` and `width`.
data_format (`str` or `ChannelDimension`, *optional*):
The channel dimension format of the image. If not provided, it will be the same as the input image.
"""
size = get_size_dict(size)
if "height" not in size or "width" not in size:
raise ValueError(f"The `size` parameter must contain the keys (height, width). Got {size.keys()}")
return center_crop(image, size=(size["height"], size["width"]), data_format=data_format, **kwargs)
def preprocess(
self,
images: ImageInput,

25
src/transformers/models/vivit/image_processing_vivit.py
View File

@ -22,7 +22,6 @@ from transformers.utils.generic import TensorType
from ...image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
from ...image_transforms import (
center_crop,
get_resize_output_image_size,
rescale,
resize,
@ -168,30 +167,6 @@ class VivitImageProcessor(BaseImageProcessor):
raise ValueError(f"Size must have 'height' and 'width' or 'shortest_edge' as keys. Got {size.keys()}")
return resize(image, size=output_size, resample=resample, data_format=data_format, **kwargs)
def center_crop(
self,
image: np.ndarray,
size: Dict[str, int],
data_format: Optional[Union[str, ChannelDimension]] = None,
**kwargs,
) -> np.ndarray:
"""
Center crop an image to `(size["height"], size["width"])`. If the input size is smaller than `size` along any
edge, the image is padded with 0's and then center cropped.
Args:
image (`np.ndarray`):
Image to center crop.
size (`Dict[str, int]`):
Size of the output image.
data_format (`str` or `ChannelDimension`, *optional*):
The channel dimension format of the image. If not provided, it will be the same as the input image.
"""
size = get_size_dict(size)
if "height" not in size or "width" not in size:
raise ValueError(f"Size must have 'height' and 'width' as keys. Got {size.keys()}")
return center_crop(image, size=(size["height"], size["width"]), data_format=data_format, **kwargs)
def rescale(
self,
image: np.ndarray,