注意
点击 此处 下载完整的示例代码
空间变换网络教程¶
作者: Ghassen HAMROUNI
在本教程中，您将学习如何使用称为空间变换网络的视觉注意力机制来增强您的网络。您可以在 DeepMind 论文 中阅读有关空间变换网络的更多信息
空间变换网络是可微注意力对任何空间变换的泛化。空间变换网络（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，它可以裁剪感兴趣区域，缩放和校正图像的方向。它可能是一种有用的机制，因为 CNN 对旋转和缩放以及更一般的仿射变换不具有不变性。
STN 最好的方面之一是能够将其简单地插入任何现有的 CNN 中，而无需进行太多修改。
# License: BSD
# Author: Ghassen Hamrouni
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
<contextlib.ExitStack object at 0x7f8530510790>
加载数据¶
在这篇文章中，我们使用经典的 MNIST 数据集进行实验。使用一个标准的卷积神经网络并增强空间变换网络。
from six.moves import urllib
opener = urllib.request.build_opener()
opener.addheaders = [('User-agent', 'Mozilla/5.0')]
urllib.request.install_opener(opener)
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)
Downloading http://yann.lecun.com/exdb/mnist/train-images-idx3-ubyte.gz
Failed to download (trying next):
<urlopen error [SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: certificate has expired (_ssl.c:1007)>
Downloading https://ossci-datasets.s3.amazonaws.com/mnist/train-images-idx3-ubyte.gz
Downloading https://ossci-datasets.s3.amazonaws.com/mnist/train-images-idx3-ubyte.gz to ./MNIST/raw/train-images-idx3-ubyte.gz
0%| | 0.00/9.91M [00:00<?, ?B/s]
100%|##########| 9.91M/9.91M [00:00<00:00, 113MB/s]
Extracting ./MNIST/raw/train-images-idx3-ubyte.gz to ./MNIST/raw
Downloading http://yann.lecun.com/exdb/mnist/train-labels-idx1-ubyte.gz
Failed to download (trying next):
<urlopen error [SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: certificate has expired (_ssl.c:1007)>
Downloading https://ossci-datasets.s3.amazonaws.com/mnist/train-labels-idx1-ubyte.gz
Downloading https://ossci-datasets.s3.amazonaws.com/mnist/train-labels-idx1-ubyte.gz to ./MNIST/raw/train-labels-idx1-ubyte.gz
0%| | 0.00/28.9k [00:00<?, ?B/s]
100%|##########| 28.9k/28.9k [00:00<00:00, 16.1MB/s]
Extracting ./MNIST/raw/train-labels-idx1-ubyte.gz to ./MNIST/raw
Downloading http://yann.lecun.com/exdb/mnist/t10k-images-idx3-ubyte.gz
Failed to download (trying next):
<urlopen error [SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: certificate has expired (_ssl.c:1007)>
Downloading https://ossci-datasets.s3.amazonaws.com/mnist/t10k-images-idx3-ubyte.gz
Downloading https://ossci-datasets.s3.amazonaws.com/mnist/t10k-images-idx3-ubyte.gz to ./MNIST/raw/t10k-images-idx3-ubyte.gz
0%| | 0.00/1.65M [00:00<?, ?B/s]
100%|##########| 1.65M/1.65M [00:00<00:00, 68.9MB/s]
Extracting ./MNIST/raw/t10k-images-idx3-ubyte.gz to ./MNIST/raw
Downloading http://yann.lecun.com/exdb/mnist/t10k-labels-idx1-ubyte.gz
Failed to download (trying next):
<urlopen error [SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: certificate has expired (_ssl.c:1007)>
Downloading https://ossci-datasets.s3.amazonaws.com/mnist/t10k-labels-idx1-ubyte.gz
Downloading https://ossci-datasets.s3.amazonaws.com/mnist/t10k-labels-idx1-ubyte.gz to ./MNIST/raw/t10k-labels-idx1-ubyte.gz
0%| | 0.00/4.54k [00:00<?, ?B/s]
100%|##########| 4.54k/4.54k [00:00<00:00, 3.08MB/s]
Extracting ./MNIST/raw/t10k-labels-idx1-ubyte.gz to ./MNIST/raw
描绘空间变换网络¶
空间变换网络归结为三个主要组件
定位网络是一个常规的 CNN，它回归变换参数。变换从未从这个数据集中明确学习，而是网络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一个坐标网格，对应于输出图像中的每个像素。
采样器使用变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 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 算法来训练模型。网络以监督的方式学习分类任务。同时，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 the STN 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，我们将检查我们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，以便在训练期间可视化变换。
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()
/opt/conda/envs/py_3.10/lib/python3.10/site-packages/torch/nn/functional.py:4969: UserWarning:
Default grid_sample and affine_grid behavior has changed to align_corners=False since 1.3.0. Please specify align_corners=True if the old behavior is desired. See the documentation of grid_sample for details.
/opt/conda/envs/py_3.10/lib/python3.10/site-packages/torch/nn/functional.py:4902: UserWarning:
Default grid_sample and affine_grid behavior has changed to align_corners=False since 1.3.0. Please specify align_corners=True if the old behavior is desired. See the documentation of grid_sample for details.
Train Epoch: 1 [0/60000 (0%)] Loss: 2.315648
Train Epoch: 1 [32000/60000 (53%)] Loss: 1.083137
/opt/conda/envs/py_3.10/lib/python3.10/site-packages/torch/nn/_reduction.py:51: UserWarning:
size_average and reduce args will be deprecated, please use reduction='sum' instead.
Test set: Average loss: 0.3183, Accuracy: 9055/10000 (91%)
Train Epoch: 2 [0/60000 (0%)] Loss: 0.658948
Train Epoch: 2 [32000/60000 (53%)] Loss: 0.328154
Test set: Average loss: 0.1462, Accuracy: 9580/10000 (96%)
Train Epoch: 3 [0/60000 (0%)] Loss: 0.331469
Train Epoch: 3 [32000/60000 (53%)] Loss: 0.207296
Test set: Average loss: 0.1242, Accuracy: 9608/10000 (96%)
Train Epoch: 4 [0/60000 (0%)] Loss: 0.418664
Train Epoch: 4 [32000/60000 (53%)] Loss: 0.110284
Test set: Average loss: 0.1021, Accuracy: 9698/10000 (97%)
Train Epoch: 5 [0/60000 (0%)] Loss: 0.214933
Train Epoch: 5 [32000/60000 (53%)] Loss: 0.245154
Test set: Average loss: 0.1872, Accuracy: 9447/10000 (94%)
Train Epoch: 6 [0/60000 (0%)] Loss: 0.388222
Train Epoch: 6 [32000/60000 (53%)] Loss: 0.098923
Test set: Average loss: 0.0697, Accuracy: 9787/10000 (98%)
Train Epoch: 7 [0/60000 (0%)] Loss: 0.083234
Train Epoch: 7 [32000/60000 (53%)] Loss: 0.179489
Test set: Average loss: 0.0635, Accuracy: 9807/10000 (98%)
Train Epoch: 8 [0/60000 (0%)] Loss: 0.182966
Train Epoch: 8 [32000/60000 (53%)] Loss: 0.056157
Test set: Average loss: 0.0570, Accuracy: 9824/10000 (98%)
Train Epoch: 9 [0/60000 (0%)] Loss: 0.167948
Train Epoch: 9 [32000/60000 (53%)] Loss: 0.096536
Test set: Average loss: 0.0610, Accuracy: 9818/10000 (98%)
Train Epoch: 10 [0/60000 (0%)] Loss: 0.127176
Train Epoch: 10 [32000/60000 (53%)] Loss: 0.199562
Test set: Average loss: 0.0670, Accuracy: 9786/10000 (98%)
Train Epoch: 11 [0/60000 (0%)] Loss: 0.167220
Train Epoch: 11 [32000/60000 (53%)] Loss: 0.073263
Test set: Average loss: 0.0547, Accuracy: 9832/10000 (98%)
Train Epoch: 12 [0/60000 (0%)] Loss: 0.134165
Train Epoch: 12 [32000/60000 (53%)] Loss: 0.128090
Test set: Average loss: 0.0559, Accuracy: 9829/10000 (98%)
Train Epoch: 13 [0/60000 (0%)] Loss: 0.112401
Train Epoch: 13 [32000/60000 (53%)] Loss: 0.161664
Test set: Average loss: 0.0529, Accuracy: 9850/10000 (98%)
Train Epoch: 14 [0/60000 (0%)] Loss: 0.095081
Train Epoch: 14 [32000/60000 (53%)] Loss: 0.164523
Test set: Average loss: 0.0532, Accuracy: 9848/10000 (98%)
Train Epoch: 15 [0/60000 (0%)] Loss: 0.101816
Train Epoch: 15 [32000/60000 (53%)] Loss: 0.104555
Test set: Average loss: 0.0506, Accuracy: 9840/10000 (98%)
Train Epoch: 16 [0/60000 (0%)] Loss: 0.114938
Train Epoch: 16 [32000/60000 (53%)] Loss: 0.151223
Test set: Average loss: 0.0510, Accuracy: 9859/10000 (99%)
Train Epoch: 17 [0/60000 (0%)] Loss: 0.219327
Train Epoch: 17 [32000/60000 (53%)] Loss: 0.183347
Test set: Average loss: 0.0753, Accuracy: 9760/10000 (98%)
Train Epoch: 18 [0/60000 (0%)] Loss: 0.085348
Train Epoch: 18 [32000/60000 (53%)] Loss: 0.128113
Test set: Average loss: 0.0465, Accuracy: 9867/10000 (99%)
Train Epoch: 19 [0/60000 (0%)] Loss: 0.067525
Train Epoch: 19 [32000/60000 (53%)] Loss: 0.130051
Test set: Average loss: 0.0441, Accuracy: 9870/10000 (99%)
Train Epoch: 20 [0/60000 (0%)] Loss: 0.050642
Train Epoch: 20 [32000/60000 (53%)] Loss: 0.024334
Test set: Average loss: 0.0561, Accuracy: 9839/10000 (98%)
脚本总运行时间：（2 分 8.860 秒）
