如何让只有一个 CNN 和一个 Pooling 层的单张图片焕发活力？答案

【问题标题】：How can I vitalize single image with only One CNN and one Pooling layers?如何让只有一个 CNN 和一个 Pooling 层的单张图片焕发活力？
【发布时间】：2022-11-18 08:26:10
【问题描述】：

我编写此示例代码以在将其传递给我的模型后仅显示单个图像该模型应该只有一个 CNN + 一个池层。或者以其他方式，我如何通过将其分配给具有一个 cnn 和一个池层的样本神经网络来 vitulize 单个图像。

`import torch
import torch.nn as nn #creating neural network
from PIL import Image
from numpy import asarray

#Set up GPU
device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
Her I am loading my image 
# load the image
image = Image.open('./img.png')
# convert image to numpy array
data = asarray(image)
print(type(data))
print(data.shape)`

现在正在建造拱门。

```
class ConvNet(nn.Module):
    def __init__(self):
        super().__init__()
        #convolutional layer
        self.layer = nn.Sequential(
            nn.Conv2d(in_channels=3, out_channels=3, kernel_size=2, stride=1, padding=0),
            nn.MaxPool2d(kernel_size=2, stride=2))
        
    def forward(self, x):
        out = self.layer(x)
        return out


convnet = ConvNet().to(device) #set up for GPU if available
convnet`
```
pass image to my model 
`outputs = convnet(data)
imshow(outputs)`
```

got the error below 

``
TypeError                                 Traceback (most recent call last)
~\AppData\Local\Temp/ipykernel_3184/1768392595.py in <module>
----> 1 outputs = convnet(data)
      2 imshow(outputs)


TypeError: conv2d() received an invalid combination of arguments - got (numpy.ndarray, Parameter, Parameter, tuple, tuple, tuple, int), but expected one of:
 * (Tensor input, Tensor weight, Tensor bias, tuple of ints stride, tuple of ints padding, tuple of ints dilation, int groups)
      didn't match because some of the arguments have invalid types: (numpy.ndarray, Parameter, Parameter, tuple, tuple, tuple, int)
 * (Tensor input, Tensor weight, Tensor bias, tuple of ints stride, str padding, tuple of ints dilation, int groups)
      didn't match because some of the arguments have invalid types: (numpy.ndarray, Parameter, Parameter, tuple, tuple, tuple, int)
``


I expect to show image after passed during this sample network

【问题讨论】：

标签： python deep-learning pytorch conv-neural-network

【解决方案1】：

CNN 的输入需要是 torch.Tensor 类型。您可以通过直接在 PIL 图像上应用变换来执行此操作，如下所示：

data = torchvision.transforms.functional.to_tensor(image)

或者

transform = torchvision.transforms.ToTensor() # can be composed with other transforms if necessary
data = transform(image)

【讨论】：

【解决方案2】：

正如 GoodDeeds 所提到的，CNN 希望数据的类型为 Tensor 您已经使用 PIL 读取了图像，然后将其转换为 NumPy 数组，您需要使用 torch.from_numpy(data) 将 NumPy 数组转换为 Tensor

下面的代码将解决这个问题

import torch
import torch.nn as nn #creating neural network
from PIL import Image
from numpy import asarray

#Set up GPU
device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
Her I am loading my image 
# load the image
image = Image.open('./img.png')
# convert image to numpy array
data = asarray(image)
data=torch.from_numpy(data)
print(type(data))
print(data.shape)`

【讨论】：