*Memos:
- My post explains how to convert and scale a PIL Image to an Image in PyTorch.
- My post explains Compose().
- My post explains ToImage().
-
My post explains ToDtype() about
scale=True. - My post explains ToTensor().
- My post explains ToPILImage() about no arguments.
- My post explains OxfordIIITPet().
PILToTensor() can convert a PIL(Pillow library) Image([H, W, C]), Image([..., C, H, W]) to a tensor([C, H, W]) and doesn't scale its values to [0.0, 1.0] as shown below:
*Memos:
- The 1st argument is
img(Required-Type:PIL Image, Image ortensor/ndarray(int/float/complex/bool)): *Memos:- A tensor can be 0D or more D.
- A ndarray can be 0D or more D.
- Don't use
img=.
-
v2is recommended to use according to V1 or V2? Which one should I use?.
from torchvision.datasets import OxfordIIITPet
from torchvision.transforms.v2 import ToImage, PILToTensor
import torch
import numpy as np
PILToTensor()
# PILToTensor()
PILImage_data = OxfordIIITPet(
root="data",
transform=None
)
Image_data = OxfordIIITPet(
root="data",
transform=ToImage()
)
Tensor_data = OxfordIIITPet(
root="data",
transform=PILToTensor()
)
Tensor_data
# Dataset OxfordIIITPet
# Number of datapoints: 3680
# Root location: data
# StandardTransform
# Transform: PILToTensor()
Tensor_data[0]
# (tensor([[[37, 35, 36, ..., 247, 249, 249],
# [35, 35, 37, ..., 246, 248, 249],
# ...,
# [28, 28, 27, ..., 59, 65, 76]],
# [[20, 18, 19, ..., 248, 248, 248],
# [18, 18, 20, ..., 247, 247, 248],
# ...,
# [27, 27, 27, ..., 94, 106, 117]],
# [[12, 10, 11, ..., 253, 253, 253],
# [10, 10, 12, ..., 251, 252, 253],
# ...,
# [35, 35, 35, ..., 214, 232, 223]]], dtype=torch.uint8), 0)
Tensor_data[0][0].size()
# torch.Size([3, 500, 394])
Tensor_data[0][0]
# tensor([[[37, 35, 36, ..., 247, 249, 249],
# [35, 35, 37, ..., 246, 248, 249],
# ...,
# [28, 28, 27, ..., 59, 65, 76]],
# [[20, 18, 19, ..., 248, 248, 248],
# [18, 18, 20, ..., 247, 247, 248],
# ...,
# [27, 27, 27, ..., 94, 106, 117]],
# [[12, 10, 11, ..., 253, 253, 253],
# [10, 10, 12, ..., 251, 252, 253],
# ...,
# [35, 35, 35, ..., 214, 232, 223]]], dtype=torch.uint8)
Tensor_data[0][1]
# 0
import matplotlib.pyplot as plt
plt.imshow(X=Tensor_data[0][0])
# TypeError: Invalid shape (3, 500, 394) for image data
ptt = PILToTensor()
ptt(PILImage_data) # It's still PIL Image.
# Dataset OxfordIIITPet
# Number of datapoints: 3680
# Root location: data
ptt(PILImage_data[0])
# (tensor([[[37, 35, 36, ..., 247, 249, 249],
# [35, 35, 37, ..., 246, 248, 249],
# ...,
# [28, 28, 27, ..., 59, 65, 76]],
# [[20, 18, 19, ..., 248, 248, 248],
# [18, 18, 20, ..., 247, 247, 248],
# ...,
# [27, 27, 27, ..., 94, 106, 117]],
# [[12, 10, 11, ..., 253, 253, 253],
# [10, 10, 12, ..., 251, 252, 253],
# ...,
# [35, 35, 35, ..., 214, 232, 223]]], dtype=torch.uint8), 0)
ptt(PILImage_data[0][0])
# tensor([[[37, 35, 36, ..., 247, 249, 249],
# [35, 35, 37, ..., 246, 248, 249],
# ...,
# [28, 28, 27, ..., 59, 65, 76]],
# [[20, 18, 19, ..., 248, 248, 248],
# [18, 18, 20, ..., 247, 247, 248],
# ...,
# [27, 27, 27, ..., 94, 106, 117]],
# [[12, 10, 11, ..., 253, 253, 253],
# [10, 10, 12, ..., 251, 252, 253],
# ...,
# [35, 35, 35, ..., 214, 232, 223]]], dtype=torch.uint8)
ptt(Image_data[0][0])
# Image([[[37, 35, 36, ..., 247, 249, 249],
# [35, 35, 37, ..., 246, 248, 249],
# ...,
# [28, 28, 27, ..., 59, 65, 76]],
# [[20, 18, 19, ..., 248, 248, 248],
# [18, 18, 20, ..., 247, 247, 248],
# ...,
# [27, 27, 27, ..., 94, 106, 117]],
# [[12, 10, 11, ..., 253, 253, 253],
# [10, 10, 12, ..., 251, 252, 253],
# ...,
# [35, 35, 35, ..., 214, 232, 223]]], dtype=torch.uint8,)
plt.imshow(X=ptt(PILImage_data[0][0]))
plt.imshow(X=ptt(Image_data[0][0]))
# TypeError: Invalid shape (3, 500, 394) for image data
ptt((torch.tensor(3), 0)) # int64
ptt((torch.tensor(3, dtype=torch.int64), 0))
# (tensor(3), 0)
ptt(torch.tensor(3))
# tensor(3)
ptt((torch.tensor([0, 1, 2, 3]), 0))
# (tensor([0, 1, 2, 3]), 0)
ptt(torch.tensor([0, 1, 2, 3]))
# tensor([0, 1, 2, 3])
ptt((torch.tensor([[0, 1, 2, 3]]), 0))
# (tensor([[0, 1, 2, 3]]), 0)
ptt(torch.tensor([[0, 1, 2, 3]]))
# tensor([[0, 1, 2, 3]])
ptt((torch.tensor([[[0, 1, 2, 3]]]), 0))
# (tensor([[[0, 1, 2, 3]]]), 0)
ptt(torch.tensor([[[0, 1, 2, 3]]]))
# tensor([[[0, 1, 2, 3]]])
ptt((torch.tensor([[[[0, 1, 2, 3]]]]), 0))
# (tensor([[[[0, 1, 2, 3]]]]), 0)
ptt(torch.tensor([[[[0, 1, 2, 3]]]]))
# tensor([[[[0, 1, 2, 3]]]])
ptt((torch.tensor([[[[[0, 1, 2, 3]]]]]), 0))
# (tensor([[[[[0, 1, 2, 3]]]]]), 0)
ptt(torch.tensor([[[[[0, 1, 2, 3]]]]]))
# tensor([[[[[0, 1, 2, 3]]]]])
ptt((torch.tensor([[0, 1, 2, 3]], dtype=torch.int32), 0))
# (tensor([[0, 1, 2, 3]], dtype=torch.int32), 0)
ptt((torch.tensor([[0, 1, 2, 3]], dtype=torch.uint8), 0))
# (tensor([[0, 1, 2, 3]], dtype=torch.uint8), 0)
ptt((torch.tensor([[0., 1., 2., 3.]]), 0)) # float32
ptt((torch.tensor([[0., 1., 2., 3.]], dtype=torch.float32), 0))
# (tensor([[0., 1., 2., 3.]]), 0)
ptt((torch.tensor([[0., 1., 2., 3.]], dtype=torch.float64), 0))
# (tensor([[0., 1., 2., 3.]], dtype=torch.float64), 0)
ptt((torch.tensor([[0.+0.j, 1.+0.j, 2.+0.j, 3.+0.j]]), 0)) # complex64
ptt((torch.tensor([[0.+0.j, 1.+0.j, 2.+0.j, 3.+0.j]],
dtype=torch.complex64), 0))
# (tensor([[0.+0.j, 1.+0.j, 2.+0.j, 3.+0.j]]), 0)
ptt((torch.tensor([[0.+0.j, 1.+0.j, 2.+0.j, 3.+0.j]],
dtype=torch.complex32), 0))
# (tensor([[0.+0.j, 1.+0.j, 2.+0.j, 3.+0.j]], dtype=torch.complex32), 0)
ptt((torch.tensor([[True, False, True, False]]), 0)) # bool
ptt((torch.tensor([[True, False, True, False]], dtype=torch.bool), 0))
# (tensor([[True, False, True, False]]), 0)
ptt((np.array(3), 0)) # int32
ptt((np.array(3, dtype=np.int32), 0))
# (array(3), 0)
ptt(np.array(3))
# array(3)
ptt((np.array([0, 1, 2, 3], dtype=np.int32), 0))
# (array([0, 1, 2, 3]), 0)
ptt(np.array([0, 1, 2, 3], dtype=np.int32))
# array([0, 1, 2, 3])
ptt((np.array([[0, 1, 2, 3]], dtype=np.int32), 0))
# (array([[0, 1, 2, 3]]), 0)
ptt(np.array([[0, 1, 2, 3]], dtype=np.int32))
# array([[0, 1, 2, 3]])
ptt((np.array([[[0, 1, 2, 3]]]), 0))
# (array([[[0, 1, 2, 3]]]), 0)
ptt(np.array([[[0, 1, 2, 3]]]))
# array([[[0, 1, 2, 3]]])
ptt((np.array([[[[0, 1, 2, 3]]]]), 0))
# (array([[[[0, 1, 2, 3]]]]), 0)
ptt(np.array([[[[0, 1, 2, 3]]]]))
# array([[[[0, 1, 2, 3]]]])
ptt((np.array([[[[[0, 1, 2, 3]]]]]), 0))
# (array([[[[[0, 1, 2, 3]]]]]), 0)
ptt(np.array([[[[[0, 1, 2, 3]]]]]))
# array([[[[[0, 1, 2, 3]]]]])
ptt((np.array([[0, 1, 2, 3]], dtype=np.int64), 0))
# (array([[0, 1, 2, 3]], dtype=int64), 0)
ptt((np.array([[0, 1, 2, 3]], dtype=np.uint8), 0))
# (array([[0, 1, 2, 3]], dtype=uint8), 0)
ptt((np.array([[0, 1, 2., 3.]], dtype=np.uint16), 0))
# (array([[0, 1, 2, 3]], dtype=uint16), 0)
ptt((np.array([[0., 1., 2., 3.]]), 0)) # float64
ptt((np.array([[0., 1., 2., 3.]], dtype=np.float64), 0))
# (array([[0., 1., 2., 3.]]), 0)
ptt((np.array([[0., 1., 2., 3.]], dtype=np.float32), 0))
# (array([[0., 1., 2., 3.]], dtype=float32), 0)
ptt((np.array([[0.+0.j, 1.+0.j, 2.+0.j, 3.+0.j]]), 0)) # complex128
ptt((np.array([[0.+0.j, 1.+0.j, 2.+0.j, 3.+0.j]], dtype=np.complex128), 0))
# (array([[0.+0.j, 1.+0.j, 2.+0.j, 3.+0.j]]), 0)
ptt((np.array([[0.+0.j, 1.+0.j, 2.+0.j, 3.+0.j]], dtype=np.complex64), 0))
# (array([[0.+0.j, 1.+0.j, 2.+0.j, 3.+0.j]], dtype=complex64), 0)
ptt((np.array([[True, False, True, False]]), 0)) # bool
ptt((np.array([[True, False, True, False]], dtype=bool), 0))
# (array([[True, False, True, False]]), 0)