计算PyTorch中Conv2d的输入和输出大小以进行图像分类

bol*_*wer 1 python image convolution pytorch tensor

我想在这里运行CIFAR10图像分类PyTorch教程- http://pytorch.org/tutorials/beginner/blitz/cifar10_tutorial.html#sphx-glr-beginner-blitz-cifar10-tutorial-py

我做了一个小的更改,并且我使用了另一个数据集。我有Wikiart数据集中要按艺术家分类的图像(标签=艺术家名称)。

这是网络的代码-

class Net(nn.Module):
    def __init__(self):
        super(Net, self).__init__()
        self.conv1 = nn.Conv2d(3, 6, 5)
        self.pool = nn.MaxPool2d(2, 2)
        self.conv2 = nn.Conv2d(6, 16, 5)
        self.fc1 = nn.Linear(16*5*5, 120)
        self.fc2 = nn.Linear(120, 84)
        self.fc3 = nn.Linear(84, 10)

    def forward(self, x):
        x = self.pool(F.relu(self.conv1(x)))
        x = self.pool(F.relu(self.conv2(x)))
        x = x.view(-1, 16*5*5)
        x = F.relu(self.fc1(x))
        x = F.relu(self.fc2(x))
        x = self.fc3(x)
        return x
Run Code Online (Sandbox Code Playgroud)

然后在代码的这一部分中,我开始训练网络。

for epoch in range(2):
     running_loss = 0.0

     for i, data in enumerate(wiki_train_dataloader, 0):
        inputs, labels = data['image'], data['class']
        print(inputs.shape)
        inputs, labels = Variable(inputs), Variable(labels)

        optimizer.zero_grad()

        # forward + backward + optimize
        outputs = net(inputs)
        loss = criterion(outputs, labels)
        loss.backward()
        optimizer.step()

        # print statistics
        running_loss += loss.data[0]
        if i % 2000 == 1999:  # print every 2000 mini-batches
            print('[%d, %5d] loss: %.3f' %
              (epoch + 1, i + 1, running_loss / 2000))
            running_loss = 0.0
Run Code Online (Sandbox Code Playgroud)

该行print(inputs.shape)提供torch.Size([4, 32, 32, 3])了我的Wikiart数据集,而在带有CIFAR10的原始示例中,它显示了torch.Size([4, 3, 32, 32])

现在,我不确定如何更改Net中的Conv2d以与兼容torch.Size([4, 32, 32, 3])

我收到此错误:

RuntimeError: Given input size: (3 x 32 x 3). Calculated output size: (6 x 28 x -1). Output size is too small at /opt/conda/conda-bld/pytorch_1503965122592/work/torch/lib/THNN/generic/SpatialConvolutionMM.c:45

在读取Wikiart数据集的图像时,我将其大小调整为(32,32),这些是3通道图像。

我尝试过的事情:

1)CIFAR10教程使用了我没有使用的转换。我无法将其合并到我的代码中。

2)更改self.conv2 = nn.Conv2d(6, 16, 5)self.conv2 = nn.Conv2d(3, 6, 5)。这给了我与上述相同的错误。我只是更改此设置以查看错误消息是否更改。

非常感谢有关如何在PyTorch中计算输入和输出大小或自动调整Tensors的任何资源。我刚开始学习Torch,发现尺寸计算很复杂。

Ego*_*kin 7

您必须将输入调整为这种格式(批处理,数字通道,高度,宽度)。当前,您的格式为(B,H,W,C)(4,32,32,3),因此您需要交换第4轴和第2轴才能用(B,C,H,W)来调整数据的形状。您可以这样操作:

inputs, labels = Variable(inputs), Variable(labels)
inputs = inputs.transpose(1,3)
... the rest
Run Code Online (Sandbox Code Playgroud)


bal*_*bok 5

我知道这是一个老问题,但在处理非标准内核大小、膨胀等时,我再次偶然发现了这个问题。这是我想出的一个函数,它为我进行计算并检查给定的输出形状:

def find_settings(shape_in, shape_out, kernel_sizes, dilation_sizes, padding_sizes, stride_sizes, transpose=False):
    from itertools import product

    import torch
    from torch import nn

    import numpy as np

    # Fake input
    x_in = torch.tensor(np.random.randn(4, 1, shape_in, shape_in), dtype=torch.float)

    # Grid search through all combinations
    for kernel, dilation, padding, stride in product(kernel_sizes, dilation_sizes, padding_sizes, stride_sizes):
        # Define a layer
        if transpose:
            layer = nn.ConvTranspose2d
        else:
            layer = nn.Conv2d
        layer = layer(
                1, 1,
                (4, kernel),
                stride=(2, stride),
                padding=(2, padding),
                dilation=(2, dilation)
            )

        # Check if layer is valid for given input shape
        try:
            x_out = layer(x_in)
        except Exception:
            continue

        # Check for shape of out tensor
        result = x_out.shape[-1]

        if shape_out == result:
            print('Correct shape for:\n ker: {}\n dil: {}\n pad: {}\n str: {}\n'.format(kernel, dilation, padding, stride))
Run Code Online (Sandbox Code Playgroud)

这是它的用法示例:

transpose = True
shape_in = 128
shape_out = 1024


kernel_sizes = [3, 4, 5, 7, 9, 11]
dilation_sizes = list(range(1, 20))
padding_sizes = list(range(15))
stride_sizes = list(range(4, 16))
find_settings(shape_in, shape_out, kernel_sizes, dilation_sizes, padding_sizes, stride_sizes, transpose)
Run Code Online (Sandbox Code Playgroud)

我希望它可以帮助将来的人们解决这个问题。请注意,它不是并行的,如果有很多选择,它可以运行一段时间。