原文：https://pytorch.org/tutorials/intermediate/spatial_transformer_tutorial.html

作者： Ghassen HAMROUNI

在本教程中，您将学习如何使用称为空间变换器网络的视觉注意力机制来扩充网络。 您可以在 DeepMind 论文中详细了解空间变压器网络。

空间变换器网络是对任何空间变换的可区别关注的概括。 空间变换器网络(简称 STN）允许神经网络学习如何对输入图像执行空间变换，以增强模型的几何不变性。 例如，它可以裁剪感兴趣的区域，缩放并校正图像的方向。 这可能是一个有用的机制，因为 CNN 不会对旋转和缩放以及更一般的仿射变换保持不变。

关于 STN 的最好的事情之一就是能够将它简单地插入到任何现有的 CNN 中。

# License: BSD
## Author: Ghassen Hamrouni


from __future__ import print_function
import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim
import torchvision
from torchvision import datasets, transforms
import matplotlib.pyplot as plt
import numpy as np


plt.ion()   # interactive mode

加载数据

在本文中，我们将尝试使用经典的 MNIST 数据集。 使用标准卷积网络和空间变换器网络。

device = torch.device("cuda" if torch.cuda.is_available() else "cpu")


## Training dataset
train_loader = torch.utils.data.DataLoader(
    datasets.MNIST(root='.', train=True, download=True,
                   transform=transforms.Compose([
                       transforms.ToTensor(),
                       transforms.Normalize((0.1307,), (0.3081,))
                   ])), batch_size=64, shuffle=True, num_workers=4)
## Test dataset
test_loader = torch.utils.data.DataLoader(
    datasets.MNIST(root='.', train=False, transform=transforms.Compose([
        transforms.ToTensor(),
        transforms.Normalize((0.1307,), (0.3081,))
    ])), batch_size=64, shuffle=True, num_workers=4)

得出：

Downloading http://yann.lecun.com/exdb/mnist/train-images-idx3-ubyte.gz to ./MNIST/raw/train-images-idx3-ubyte.gz
Extracting ./MNIST/raw/train-images-idx3-ubyte.gz to ./MNIST/raw
Downloading http://yann.lecun.com/exdb/mnist/train-labels-idx1-ubyte.gz to ./MNIST/raw/train-labels-idx1-ubyte.gz
Extracting ./MNIST/raw/train-labels-idx1-ubyte.gz to ./MNIST/raw
Downloading http://yann.lecun.com/exdb/mnist/t10k-images-idx3-ubyte.gz to ./MNIST/raw/t10k-images-idx3-ubyte.gz
Extracting ./MNIST/raw/t10k-images-idx3-ubyte.gz to ./MNIST/raw
Downloading http://yann.lecun.com/exdb/mnist/t10k-labels-idx1-ubyte.gz to ./MNIST/raw/t10k-labels-idx1-ubyte.gz
Extracting ./MNIST/raw/t10k-labels-idx1-ubyte.gz to ./MNIST/raw
Processing...
Done!

描述空间变压器网络

空间变压器网络可归结为三个主要组成部分：

• 本地化网络是常规的 CNN，可以对转换参数进行回归。 永远不会从此数据集中显式学习变换，而是网络会自动学习增强全局精度的空间变换。
• 网格生成器在输入图像中生成与来自输出图像的每个像素相对应的坐标网格。
• 采样器使用转换的参数，并将其应用于输入图像。

注意：

我们需要包含 affine_grid 和 grid_sample 模块的最新版本的 PyTorch。

class Net(nn.Module):
    def __init__(self):
        super(Net, self).__init__()
        self.conv1 = nn.Conv2d(1, 10, kernel_size=5)
        self.conv2 = nn.Conv2d(10, 20, kernel_size=5)
        self.conv2_drop = nn.Dropout2d()
        self.fc1 = nn.Linear(320, 50)
        self.fc2 = nn.Linear(50, 10)


        # Spatial transformer localization-network
        self.localization = nn.Sequential(
            nn.Conv2d(1, 8, kernel_size=7),
            nn.MaxPool2d(2, stride=2),
            nn.ReLU(True),
            nn.Conv2d(8, 10, kernel_size=5),
            nn.MaxPool2d(2, stride=2),
            nn.ReLU(True)
        )


        # Regressor for the 3 * 2 affine matrix
        self.fc_loc = nn.Sequential(
            nn.Linear(10 * 3 * 3, 32),
            nn.ReLU(True),
            nn.Linear(32, 3 * 2)
        )


        # Initialize the weights/bias with identity transformation
        self.fc_loc[2].weight.data.zero_()
        self.fc_loc[2].bias.data.copy_(torch.tensor([1, 0, 0, 0, 1, 0], dtype=torch.float))


    # Spatial transformer network forward function
    def stn(self, x):
        xs = self.localization(x)
        xs = xs.view(-1, 10 * 3 * 3)
        theta = self.fc_loc(xs)
        theta = theta.view(-1, 2, 3)


        grid = F.affine_grid(theta, x.size())
        x = F.grid_sample(x, grid)


        return x


    def forward(self, x):
        # transform the input
        x = self.stn(x)


        # Perform the usual forward pass
        x = F.relu(F.max_pool2d(self.conv1(x), 2))
        x = F.relu(F.max_pool2d(self.conv2_drop(self.conv2(x)), 2))
        x = x.view(-1, 320)
        x = F.relu(self.fc1(x))
        x = F.dropout(x, training=self.training)
        x = self.fc2(x)
        return F.log_softmax(x, dim=1)


model = Net().to(device)

训练模型

现在，让我们使用 SGD 算法训练模型。 网络正在以监督方式学习分类任务。 同时，该模型以端到端的方式自动学习 STN。

optimizer = optim.SGD(model.parameters(), lr=0.01)


def train(epoch):
    model.train()
    for batch_idx, (data, target) in enumerate(train_loader):
        data, target = data.to(device), target.to(device)


        optimizer.zero_grad()
        output = model(data)
        loss = F.nll_loss(output, target)
        loss.backward()
        optimizer.step()
        if batch_idx % 500 == 0:
            print('Train Epoch: {} [{}/{} ({:.0f}%)]\tLoss: {:.6f}'.format(
                epoch, batch_idx * len(data), len(train_loader.dataset),
                100\. * batch_idx / len(train_loader), loss.item()))
## ## A simple test procedure to measure STN the performances on MNIST.
## 
def test():
    with torch.no_grad():
        model.eval()
        test_loss = 0
        correct = 0
        for data, target in test_loader:
            data, target = data.to(device), target.to(device)
            output = model(data)


            # sum up batch loss
            test_loss += F.nll_loss(output, target, size_average=False).item()
            # get the index of the max log-probability
            pred = output.max(1, keepdim=True)[1]
            correct += pred.eq(target.view_as(pred)).sum().item()


        test_loss /= len(test_loader.dataset)
        print('\nTest set: Average loss: {:.4f}, Accuracy: {}/{} ({:.0f}%)\n'
              .format(test_loss, correct, len(test_loader.dataset),
                      100\. * correct / len(test_loader.dataset)))

可视化 STN 结果

现在，我们将检查学习到的视觉注意力机制的结果。

我们定义了一个小的辅助函数，以便在训练时可视化转换。

def convert_image_np(inp):
    """Convert a Tensor to numpy image."""
    inp = inp.numpy().transpose((1, 2, 0))
    mean = np.array([0.485, 0.456, 0.406])
    std = np.array([0.229, 0.224, 0.225])
    inp = std * inp + mean
    inp = np.clip(inp, 0, 1)
    return inp


## We want to visualize the output of the spatial transformers layer
## after the training, we visualize a batch of input images and
## the corresponding transformed batch using STN.


def visualize_stn():
    with torch.no_grad():
        # Get a batch of training data
        data = next(iter(test_loader))[0].to(device)


        input_tensor = data.cpu()
        transformed_input_tensor = model.stn(data).cpu()


        in_grid = convert_image_np(
            torchvision.utils.make_grid(input_tensor))


        out_grid = convert_image_np(
            torchvision.utils.make_grid(transformed_input_tensor))


        # Plot the results side-by-side
        f, axarr = plt.subplots(1, 2)
        axarr[0].imshow(in_grid)
        axarr[0].set_title('Dataset Images')


        axarr[1].imshow(out_grid)
        axarr[1].set_title('Transformed Images')


for epoch in range(1, 20 + 1):
    train(epoch)
    test()


## Visualize the STN transformation on some input batch
visualize_stn()


plt.ioff()
plt.show()

得出:

Train Epoch: 1 [0/60000 (0%)]   Loss: 2.312544
Train Epoch: 1 [32000/60000 (53%)]      Loss: 0.865688


Test set: Average loss: 0.2105, Accuracy: 9426/10000 (94%)


Train Epoch: 2 [0/60000 (0%)]   Loss: 0.528199
Train Epoch: 2 [32000/60000 (53%)]      Loss: 0.273284


Test set: Average loss: 0.1150, Accuracy: 9661/10000 (97%)


Train Epoch: 3 [0/60000 (0%)]   Loss: 0.312562
Train Epoch: 3 [32000/60000 (53%)]      Loss: 0.496166


Test set: Average loss: 0.1130, Accuracy: 9661/10000 (97%)


Train Epoch: 4 [0/60000 (0%)]   Loss: 0.346181
Train Epoch: 4 [32000/60000 (53%)]      Loss: 0.206084


Test set: Average loss: 0.0875, Accuracy: 9730/10000 (97%)


Train Epoch: 5 [0/60000 (0%)]   Loss: 0.351175
Train Epoch: 5 [32000/60000 (53%)]      Loss: 0.388225


Test set: Average loss: 0.0659, Accuracy: 9802/10000 (98%)


Train Epoch: 6 [0/60000 (0%)]   Loss: 0.122667
Train Epoch: 6 [32000/60000 (53%)]      Loss: 0.258372


Test set: Average loss: 0.0791, Accuracy: 9759/10000 (98%)


Train Epoch: 7 [0/60000 (0%)]   Loss: 0.190197
Train Epoch: 7 [32000/60000 (53%)]      Loss: 0.154468


Test set: Average loss: 0.0647, Accuracy: 9791/10000 (98%)


Train Epoch: 8 [0/60000 (0%)]   Loss: 0.121149
Train Epoch: 8 [32000/60000 (53%)]      Loss: 0.288490


Test set: Average loss: 0.0583, Accuracy: 9821/10000 (98%)


Train Epoch: 9 [0/60000 (0%)]   Loss: 0.244609
Train Epoch: 9 [32000/60000 (53%)]      Loss: 0.023396


Test set: Average loss: 0.0685, Accuracy: 9778/10000 (98%)


Train Epoch: 10 [0/60000 (0%)]  Loss: 0.256878
Train Epoch: 10 [32000/60000 (53%)]     Loss: 0.091626


Test set: Average loss: 0.0684, Accuracy: 9783/10000 (98%)


Train Epoch: 11 [0/60000 (0%)]  Loss: 0.181910
Train Epoch: 11 [32000/60000 (53%)]     Loss: 0.113193


Test set: Average loss: 0.0492, Accuracy: 9856/10000 (99%)


Train Epoch: 12 [0/60000 (0%)]  Loss: 0.081072
Train Epoch: 12 [32000/60000 (53%)]     Loss: 0.082513


Test set: Average loss: 0.0670, Accuracy: 9800/10000 (98%)


Train Epoch: 13 [0/60000 (0%)]  Loss: 0.180748
Train Epoch: 13 [32000/60000 (53%)]     Loss: 0.194512


Test set: Average loss: 0.0439, Accuracy: 9874/10000 (99%)


Train Epoch: 14 [0/60000 (0%)]  Loss: 0.099560
Train Epoch: 14 [32000/60000 (53%)]     Loss: 0.084377


Test set: Average loss: 0.0416, Accuracy: 9880/10000 (99%)


Train Epoch: 15 [0/60000 (0%)]  Loss: 0.070021
Train Epoch: 15 [32000/60000 (53%)]     Loss: 0.241336


Test set: Average loss: 0.0588, Accuracy: 9820/10000 (98%)


Train Epoch: 16 [0/60000 (0%)]  Loss: 0.060536
Train Epoch: 16 [32000/60000 (53%)]     Loss: 0.053016


Test set: Average loss: 0.0405, Accuracy: 9877/10000 (99%)


Train Epoch: 17 [0/60000 (0%)]  Loss: 0.207369
Train Epoch: 17 [32000/60000 (53%)]     Loss: 0.069607


Test set: Average loss: 0.1006, Accuracy: 9685/10000 (97%)


Train Epoch: 18 [0/60000 (0%)]  Loss: 0.127503
Train Epoch: 18 [32000/60000 (53%)]     Loss: 0.070724


Test set: Average loss: 0.0659, Accuracy: 9814/10000 (98%)


Train Epoch: 19 [0/60000 (0%)]  Loss: 0.176861
Train Epoch: 19 [32000/60000 (53%)]     Loss: 0.116980


Test set: Average loss: 0.0413, Accuracy: 9871/10000 (99%)


Train Epoch: 20 [0/60000 (0%)]  Loss: 0.146933
Train Epoch: 20 [32000/60000 (53%)]     Loss: 0.245741


Test set: Average loss: 0.0346, Accuracy: 9892/10000 (99%)

脚本的总运行时间：(2 分钟 3.339 秒）

Download Python source code: spatial_transformer_tutorial.py Download Jupyter notebook: spatial_transformer_tutorial.ipynb

由狮身人面像画廊生成的画廊