Shortcuts

Spatial Transformer Networks Tutorial

Author: Ghassen HAMROUNI

../_images/FSeq.png

In this tutorial, you will learn how to augment your network using a visual attention mechanism called spatial transformer networks. You can read more about the spatial transformer networks in the DeepMind paper

Spatial transformer networks are a generalization of differentiable attention to any spatial transformation. Spatial transformer networks (STN for short) allow a neural network to learn how to perform spatial transformations on the input image in order to enhance the geometric invariance of the model. For example, it can crop a region of interest, scale and correct the orientation of an image. It can be a useful mechanism because CNNs are not invariant to rotation and scale and more general affine transformations.

One of the best things about STN is the ability to simply plug it into any existing CNN with very little modification.

# 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 0x7f1ccf384a30>

Loading the data

In this post we experiment with the classic MNIST dataset. Using a standard convolutional network augmented with a spatial transformer network.

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
Downloading http://yann.lecun.com/exdb/mnist/train-images-idx3-ubyte.gz to ./MNIST/raw/train-images-idx3-ubyte.gz

  0%|          | 0/9912422 [00:00<?, ?it/s]
100%|##########| 9912422/9912422 [00:00<00:00, 244041906.07it/s]
Extracting ./MNIST/raw/train-images-idx3-ubyte.gz to ./MNIST/raw

Downloading http://yann.lecun.com/exdb/mnist/train-labels-idx1-ubyte.gz
Downloading http://yann.lecun.com/exdb/mnist/train-labels-idx1-ubyte.gz to ./MNIST/raw/train-labels-idx1-ubyte.gz

  0%|          | 0/28881 [00:00<?, ?it/s]
100%|##########| 28881/28881 [00:00<00:00, 91908720.66it/s]
Extracting ./MNIST/raw/train-labels-idx1-ubyte.gz to ./MNIST/raw

Downloading http://yann.lecun.com/exdb/mnist/t10k-images-idx3-ubyte.gz
Downloading http://yann.lecun.com/exdb/mnist/t10k-images-idx3-ubyte.gz to ./MNIST/raw/t10k-images-idx3-ubyte.gz

  0%|          | 0/1648877 [00:00<?, ?it/s]
100%|##########| 1648877/1648877 [00:00<00:00, 106734954.81it/s]
Extracting ./MNIST/raw/t10k-images-idx3-ubyte.gz to ./MNIST/raw

Downloading http://yann.lecun.com/exdb/mnist/t10k-labels-idx1-ubyte.gz
Downloading http://yann.lecun.com/exdb/mnist/t10k-labels-idx1-ubyte.gz to ./MNIST/raw/t10k-labels-idx1-ubyte.gz

  0%|          | 0/4542 [00:00<?, ?it/s]
100%|##########| 4542/4542 [00:00<00:00, 28647411.68it/s]
Extracting ./MNIST/raw/t10k-labels-idx1-ubyte.gz to ./MNIST/raw

Depicting spatial transformer networks

Spatial transformer networks boils down to three main components :

  • The localization network is a regular CNN which regresses the transformation parameters. The transformation is never learned explicitly from this dataset, instead the network learns automatically the spatial transformations that enhances the global accuracy.

  • The grid generator generates a grid of coordinates in the input image corresponding to each pixel from the output image.

  • The sampler uses the parameters of the transformation and applies it to the input image.

../_images/stn-arch.png

Note

We need the latest version of PyTorch that contains affine_grid and grid_sample modules.

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)

Training the model

Now, let’s use the SGD algorithm to train the model. The network is learning the classification task in a supervised way. In the same time the model is learning STN automatically in an end-to-end fashion.

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)))

Visualizing the STN results

Now, we will inspect the results of our learned visual attention mechanism.

We define a small helper function in order to visualize the transformations while training.

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()
Dataset Images, Transformed Images
Train Epoch: 1 [0/60000 (0%)]   Loss: 2.315648
Train Epoch: 1 [32000/60000 (53%)]      Loss: 1.079918

Test set: Average loss: 0.2936, Accuracy: 9182/10000 (92%)

Train Epoch: 2 [0/60000 (0%)]   Loss: 0.598581
Train Epoch: 2 [32000/60000 (53%)]      Loss: 0.320083

Test set: Average loss: 0.1510, Accuracy: 9552/10000 (96%)

Train Epoch: 3 [0/60000 (0%)]   Loss: 0.378362
Train Epoch: 3 [32000/60000 (53%)]      Loss: 0.241178

Test set: Average loss: 0.2499, Accuracy: 9198/10000 (92%)

Train Epoch: 4 [0/60000 (0%)]   Loss: 0.657763
Train Epoch: 4 [32000/60000 (53%)]      Loss: 0.167312

Test set: Average loss: 0.1247, Accuracy: 9625/10000 (96%)

Train Epoch: 5 [0/60000 (0%)]   Loss: 0.268920
Train Epoch: 5 [32000/60000 (53%)]      Loss: 0.181915

Test set: Average loss: 0.1089, Accuracy: 9663/10000 (97%)

Train Epoch: 6 [0/60000 (0%)]   Loss: 0.161158
Train Epoch: 6 [32000/60000 (53%)]      Loss: 0.097544

Test set: Average loss: 0.0728, Accuracy: 9790/10000 (98%)

Train Epoch: 7 [0/60000 (0%)]   Loss: 0.088728
Train Epoch: 7 [32000/60000 (53%)]      Loss: 0.184994

Test set: Average loss: 0.0632, Accuracy: 9804/10000 (98%)

Train Epoch: 8 [0/60000 (0%)]   Loss: 0.216909
Train Epoch: 8 [32000/60000 (53%)]      Loss: 0.092944

Test set: Average loss: 0.0714, Accuracy: 9771/10000 (98%)

Train Epoch: 9 [0/60000 (0%)]   Loss: 0.102966
Train Epoch: 9 [32000/60000 (53%)]      Loss: 0.057972

Test set: Average loss: 0.0637, Accuracy: 9811/10000 (98%)

Train Epoch: 10 [0/60000 (0%)]  Loss: 0.098537
Train Epoch: 10 [32000/60000 (53%)]     Loss: 0.167220

Test set: Average loss: 0.0540, Accuracy: 9840/10000 (98%)

Train Epoch: 11 [0/60000 (0%)]  Loss: 0.136501
Train Epoch: 11 [32000/60000 (53%)]     Loss: 0.091753

Test set: Average loss: 0.0642, Accuracy: 9815/10000 (98%)

Train Epoch: 12 [0/60000 (0%)]  Loss: 0.120580
Train Epoch: 12 [32000/60000 (53%)]     Loss: 0.138647

Test set: Average loss: 0.0517, Accuracy: 9846/10000 (98%)

Train Epoch: 13 [0/60000 (0%)]  Loss: 0.139712
Train Epoch: 13 [32000/60000 (53%)]     Loss: 0.079615

Test set: Average loss: 0.0476, Accuracy: 9849/10000 (98%)

Train Epoch: 14 [0/60000 (0%)]  Loss: 0.103130
Train Epoch: 14 [32000/60000 (53%)]     Loss: 0.182314

Test set: Average loss: 0.0495, Accuracy: 9858/10000 (99%)

Train Epoch: 15 [0/60000 (0%)]  Loss: 0.031161
Train Epoch: 15 [32000/60000 (53%)]     Loss: 0.055238

Test set: Average loss: 0.0444, Accuracy: 9863/10000 (99%)

Train Epoch: 16 [0/60000 (0%)]  Loss: 0.048907
Train Epoch: 16 [32000/60000 (53%)]     Loss: 0.158820

Test set: Average loss: 0.0446, Accuracy: 9880/10000 (99%)

Train Epoch: 17 [0/60000 (0%)]  Loss: 0.223445
Train Epoch: 17 [32000/60000 (53%)]     Loss: 0.111553

Test set: Average loss: 0.0476, Accuracy: 9849/10000 (98%)

Train Epoch: 18 [0/60000 (0%)]  Loss: 0.059782
Train Epoch: 18 [32000/60000 (53%)]     Loss: 0.138136

Test set: Average loss: 0.0525, Accuracy: 9849/10000 (98%)

Train Epoch: 19 [0/60000 (0%)]  Loss: 0.076469
Train Epoch: 19 [32000/60000 (53%)]     Loss: 0.148273

Test set: Average loss: 0.0472, Accuracy: 9861/10000 (99%)

Train Epoch: 20 [0/60000 (0%)]  Loss: 0.131952
Train Epoch: 20 [32000/60000 (53%)]     Loss: 0.015045

Test set: Average loss: 0.0475, Accuracy: 9849/10000 (98%)

Total running time of the script: ( 2 minutes 5.041 seconds)

Gallery generated by Sphinx-Gallery

Docs

Access comprehensive developer documentation for PyTorch

View Docs

Tutorials

Get in-depth tutorials for beginners and advanced developers

View Tutorials

Resources

Find development resources and get your questions answered

View Resources