ViPer / diffusers /utils /pil_utils.py
miaw1419's picture
Upload 472 files
0aaa1f1 verified
from typing import List
import PIL.Image
import PIL.ImageOps
from packaging import version
from PIL import Image
if version.parse(version.parse(PIL.__version__).base_version) >= version.parse("9.1.0"):
PIL_INTERPOLATION = {
"linear": PIL.Image.Resampling.BILINEAR,
"bilinear": PIL.Image.Resampling.BILINEAR,
"bicubic": PIL.Image.Resampling.BICUBIC,
"lanczos": PIL.Image.Resampling.LANCZOS,
"nearest": PIL.Image.Resampling.NEAREST,
}
else:
PIL_INTERPOLATION = {
"linear": PIL.Image.LINEAR,
"bilinear": PIL.Image.BILINEAR,
"bicubic": PIL.Image.BICUBIC,
"lanczos": PIL.Image.LANCZOS,
"nearest": PIL.Image.NEAREST,
}
def pt_to_pil(images):
"""
Convert a torch image to a PIL image.
"""
images = (images / 2 + 0.5).clamp(0, 1)
images = images.cpu().permute(0, 2, 3, 1).float().numpy()
images = numpy_to_pil(images)
return images
def numpy_to_pil(images):
"""
Convert a numpy image or a batch of images to a PIL image.
"""
if images.ndim == 3:
images = images[None, ...]
images = (images * 255).round().astype("uint8")
if images.shape[-1] == 1:
# special case for grayscale (single channel) images
pil_images = [Image.fromarray(image.squeeze(), mode="L") for image in images]
else:
pil_images = [Image.fromarray(image) for image in images]
return pil_images
def make_image_grid(images: List[PIL.Image.Image], rows: int, cols: int, resize: int = None) -> PIL.Image.Image:
"""
Prepares a single grid of images. Useful for visualization purposes.
"""
assert len(images) == rows * cols
if resize is not None:
images = [img.resize((resize, resize)) for img in images]
w, h = images[0].size
grid = Image.new("RGB", size=(cols * w, rows * h))
for i, img in enumerate(images):
grid.paste(img, box=(i % cols * w, i // cols * h))
return grid