比较与torchvision.transforms.ToTensor的差异

torchvision.transforms.ToTensor

class torchvision.transforms.ToTensor

更多内容详见torchvision.transforms.ToTensor。

mindspore.dataset.vision.ToTensor

class mindspore.dataset.vision.ToTensor(
    output_type=np.float32
    )

更多内容详见mindspore.dataset.vision.ToTensor。

差异对比

PyTorch：将PIL类型的Image或Numpy数组转换为torch中的Tensor，输入的Numpy数组通常是<H, W, C>格式且取值在[0, 255]范围，输出是<C, H, W>格式且取值在[0.0, 1.0]的torch Tensor。

MindSpore：输入为PIL类型的图像或<H, W, C>格式且取值在[0, 255]范围内的Numpy数组，输出为[0.0, 1.0]范围内且具有<C, H, W>格式的Numpy数组；等同于在原始输入图像上做了通道转换及像素值归一化两种操作。

分类	子类	PyTorch	MindSpore	差异
参数	参数 1	-	output_type	指定输出Numpy数组的类型

代码示例

import numpy as np
from PIL import Image
from download import download
from torchvision import transforms
import mindspore.dataset.vision as vision

url = "https://mindspore-website.obs.cn-north-4.myhuaweicloud.com/notebook/datasets/flamingos.jpg"
download(url, './flamingos.jpg', replace=True)
img = Image.open('flamingos.jpg')

# In MindSpore, ToTensor convert PIL Image into numpy array.
to_tensor = vision.ToTensor()
img_data = to_tensor(img)
print("img_data shape:", img_data.shape)
print("img_data type:", type(img_data))
# Out:
# img_data shape: (3, 292, 471)
# img_data type: <class 'numpy.ndarray'>

# In torch, ToTensor transforms the input to tensor.
image_transform = transforms.Compose([transforms.ToTensor()])
img_data = image_transform(img)
print("img_data shape:", img_data.shape)
print("img_data type:", type(img_data))
# Out:
# img_data shape: torch.Size([3, 292, 471])
# img_data type: <class 'torch.Tensor'>