File size: 2,605 Bytes
8b306b3 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 | # Copyright (c) 2025 ByteDance Ltd. and/or its affiliates.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import math
from typing import List, Union
import torch
from PIL import Image
from torchvision.transforms import functional as TVF
from torchvision.transforms.functional import InterpolationMode, to_tensor
class AreaResize:
def __init__(
self,
max_area: float,
downsample_only: bool = False,
interpolation: InterpolationMode = InterpolationMode.BICUBIC,
):
self.max_area = max_area
self.downsample_only = downsample_only
self.interpolation = interpolation
def __call__(self, image: Union[torch.Tensor, Image.Image, List[Image.Image]]):
if isinstance(image, torch.Tensor):
height, width = image.shape[-2:]
elif isinstance(image, Image.Image):
width, height = image.size
elif isinstance(image, list) and isinstance(image[0], Image.Image):
width, height = image[0].size
else:
raise NotImplementedError
scale = math.sqrt(self.max_area / (height * width))
# keep original height and width for small pictures.
scale = 1 if scale >= 1 and self.downsample_only else scale
resized_height, resized_width = round(height * scale), round(width * scale)
if isinstance(image, list) and isinstance(image[0], Image.Image):
image = torch.stack(
[
to_tensor(
TVF.resize(
_image,
size=(resized_height, resized_width),
interpolation=self.interpolation,
)
)
for _image in image
]
)
else:
image = TVF.resize(
image,
size=(resized_height, resized_width),
interpolation=self.interpolation,
)
if isinstance(image, Image.Image):
image = to_tensor(image)
return image
|