Transfer Learning for Computer Vision Tutorial

Author: Sasank Chilamkurthy

In this tutorial, you will learn how to train a convolutional neural network for image classification using transfer learning. You can read more about the transfer learning at cs231n notes

Quoting these notes,

In practice, very few people train an entire Convolutional Network from scratch (with random initialization), because it is relatively rare to have a dataset of sufficient size. Instead, it is common to pretrain a ConvNet on a very large dataset (e.g. ImageNet, which contains 1.2 million images with 1000 categories), and then use the ConvNet either as an initialization or a fixed feature extractor for the task of interest.

These two major transfer learning scenarios look as follows:

• Finetuning the convnet: Instead of random initializaion, we initialize the network with a pretrained network, like the one that is trained on imagenet 1000 dataset. Rest of the training looks as usual.
• ConvNet as fixed feature extractor: Here, we will freeze the weights for all of the network except that of the final fully connected layer. This last fully connected layer is replaced with a new one with random weights and only this layer is trained.

# License: BSD
# Author: Sasank Chilamkurthy

from __future__ import print_function, division

import torch
import torch.nn as nn
import torch.optim as optim
from torch.optim import lr_scheduler
import numpy as np
import torchvision
from torchvision import datasets, models, transforms
import matplotlib.pyplot as plt
import time
import os
import copy

plt.ion()   # interactive mode

Load Data

We will use torchvision and torch.utils.data packages for loading the data.

The problem we’re going to solve today is to train a model to classify ants and bees. We have about 120 training images each for ants and bees. There are 75 validation images for each class. Usually, this is a very small dataset to generalize upon, if trained from scratch. Since we are using transfer learning, we should be able to generalize reasonably well.

This dataset is a very small subset of imagenet.

Note

Download the data from here and extract it to the current directory.

# Data augmentation and normalization for training
# Just normalization for validation
data_transforms = {
    'train': transforms.Compose([
        transforms.RandomResizedCrop(224),
        transforms.RandomHorizontalFlip(),
        transforms.ToTensor(),
        transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])
    ]),
    'val': transforms.Compose([
        transforms.Resize(256),
        transforms.CenterCrop(224),
        transforms.ToTensor(),
        transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])
    ]),
}

data_dir = 'data/hymenoptera_data'
image_datasets = {x: datasets.ImageFolder(os.path.join(data_dir, x),
                                          data_transforms[x])
                  for x in ['train', 'val']}
dataloaders = {x: torch.utils.data.DataLoader(image_datasets[x], batch_size=4,
                                             shuffle=True, num_workers=4)
              for x in ['train', 'val']}
dataset_sizes = {x: len(image_datasets[x]) for x in ['train', 'val']}
class_names = image_datasets['train'].classes

device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")

Visualize a few images

Let’s visualize a few training images so as to understand the data augmentations.

def imshow(inp, title=None):
    """Imshow for Tensor."""
    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)
    plt.imshow(inp)
    if title is not None:
        plt.title(title)
    plt.pause(0.001)  # pause a bit so that plots are updated


# Get a batch of training data
inputs, classes = next(iter(dataloaders['train']))

# Make a grid from batch
out = torchvision.utils.make_grid(inputs)

imshow(out, title=[class_names[x] for x in classes])

Training the model

Now, let’s write a general function to train a model. Here, we will illustrate:

• Scheduling the learning rate
• Saving the best model

In the following, parameter scheduler is an LR scheduler object from torch.optim.lr_scheduler.

def train_model(model, criterion, optimizer, scheduler, num_epochs=25):
    since = time.time()

    best_model_wts = copy.deepcopy(model.state_dict())
    best_acc = 0.0

    for epoch in range(num_epochs):
        print('Epoch {}/{}'.format(epoch, num_epochs - 1))
        print('-' * 10)

        # Each epoch has a training and validation phase
        for phase in ['train', 'val']:
            if phase == 'train':
                model.train()  # Set model to training mode
            else:
                model.eval()   # Set model to evaluate mode

            running_loss = 0.0
            running_corrects = 0

            # Iterate over data.
            for inputs, labels in dataloaders[phase]:
                inputs = inputs.to(device)
                labels = labels.to(device)

                # zero the parameter gradients
                optimizer.zero_grad()

                # forward
                # track history if only in train
                with torch.set_grad_enabled(phase == 'train'):
                    outputs = model(inputs)
                    _, preds = torch.max(outputs, 1)
                    loss = criterion(outputs, labels)

                    # backward + optimize only if in training phase
                    if phase == 'train':
                        loss.backward()
                        optimizer.step()

                # statistics
                running_loss += loss.item() * inputs.size(0)
                running_corrects += torch.sum(preds == labels.data)
            if phase == 'train':
                scheduler.step()

            epoch_loss = running_loss / dataset_sizes[phase]
            epoch_acc = running_corrects.double() / dataset_sizes[phase]

            print('{} Loss: {:.4f} Acc: {:.4f}'.format(
                phase, epoch_loss, epoch_acc))

            # deep copy the model
            if phase == 'val' and epoch_acc > best_acc:
                best_acc = epoch_acc
                best_model_wts = copy.deepcopy(model.state_dict())

        print()

    time_elapsed = time.time() - since
    print('Training complete in {:.0f}m {:.0f}s'.format(
        time_elapsed // 60, time_elapsed % 60))
    print('Best val Acc: {:4f}'.format(best_acc))

    # load best model weights
    model.load_state_dict(best_model_wts)
    return model

Visualizing the model predictions

Generic function to display predictions for a few images

def visualize_model(model, num_images=6):
    was_training = model.training
    model.eval()
    images_so_far = 0
    fig = plt.figure()

    with torch.no_grad():
        for i, (inputs, labels) in enumerate(dataloaders['val']):
            inputs = inputs.to(device)
            labels = labels.to(device)

            outputs = model(inputs)
            _, preds = torch.max(outputs, 1)

            for j in range(inputs.size()[0]):
                images_so_far += 1
                ax = plt.subplot(num_images//2, 2, images_so_far)
                ax.axis('off')
                ax.set_title('predicted: {}'.format(class_names[preds[j]]))
                imshow(inputs.cpu().data[j])

                if images_so_far == num_images:
                    model.train(mode=was_training)
                    return
        model.train(mode=was_training)

Finetuning the convnet

Load a pretrained model and reset final fully connected layer.

model_ft = models.resnet18(pretrained=True)
num_ftrs = model_ft.fc.in_features
# Here the size of each output sample is set to 2.
# Alternatively, it can be generalized to nn.Linear(num_ftrs, len(class_names)).
model_ft.fc = nn.Linear(num_ftrs, 2)

model_ft = model_ft.to(device)

criterion = nn.CrossEntropyLoss()

# Observe that all parameters are being optimized
optimizer_ft = optim.SGD(model_ft.parameters(), lr=0.001, momentum=0.9)

# Decay LR by a factor of 0.1 every 7 epochs
exp_lr_scheduler = lr_scheduler.StepLR(optimizer_ft, step_size=7, gamma=0.1)

Train and evaluate

It should take around 15-25 min on CPU. On GPU though, it takes less than a minute.

model_ft = train_model(model_ft, criterion, optimizer_ft, exp_lr_scheduler,
                       num_epochs=25)

Out:

Epoch 0/24
----------
train Loss: 0.6750 Acc: 0.6557
val Loss: 0.3182 Acc: 0.9150

Epoch 1/24
----------
train Loss: 0.9820 Acc: 0.7172
val Loss: 0.4363 Acc: 0.8627

Epoch 2/24
----------
train Loss: 0.6248 Acc: 0.7869
val Loss: 0.4056 Acc: 0.8693

Epoch 3/24
----------
train Loss: 0.5376 Acc: 0.7582
val Loss: 1.0370 Acc: 0.7386

Epoch 4/24
----------
train Loss: 0.6330 Acc: 0.7951
val Loss: 0.3141 Acc: 0.8824

Epoch 5/24
----------
train Loss: 0.5061 Acc: 0.8074
val Loss: 0.6154 Acc: 0.8366

Epoch 6/24
----------
train Loss: 0.4283 Acc: 0.7992
val Loss: 0.3225 Acc: 0.8562

Epoch 7/24
----------
train Loss: 0.3968 Acc: 0.8361
val Loss: 0.2257 Acc: 0.9150

Epoch 8/24
----------
train Loss: 0.2702 Acc: 0.8770
val Loss: 0.2366 Acc: 0.9085

Epoch 9/24
----------
train Loss: 0.3437 Acc: 0.8730
val Loss: 0.2389 Acc: 0.9216

Epoch 10/24
----------
train Loss: 0.2744 Acc: 0.8852
val Loss: 0.2374 Acc: 0.9150

Epoch 11/24
----------
train Loss: 0.3440 Acc: 0.8811
val Loss: 0.2333 Acc: 0.9216

Epoch 12/24
----------
train Loss: 0.2161 Acc: 0.9057
val Loss: 0.2339 Acc: 0.9346

Epoch 13/24
----------
train Loss: 0.2950 Acc: 0.8648
val Loss: 0.2525 Acc: 0.9281

Epoch 14/24
----------
train Loss: 0.3029 Acc: 0.8484
val Loss: 0.2462 Acc: 0.9346

Epoch 15/24
----------
train Loss: 0.3016 Acc: 0.8770
val Loss: 0.2564 Acc: 0.9150

Epoch 16/24
----------
train Loss: 0.3248 Acc: 0.8566
val Loss: 0.2462 Acc: 0.9216

Epoch 17/24
----------
train Loss: 0.2103 Acc: 0.9221
val Loss: 0.2425 Acc: 0.9281

Epoch 18/24
----------
train Loss: 0.2811 Acc: 0.8893
val Loss: 0.2421 Acc: 0.9346

Epoch 19/24
----------
train Loss: 0.2590 Acc: 0.8811
val Loss: 0.2326 Acc: 0.9216

Epoch 20/24
----------
train Loss: 0.2769 Acc: 0.8975
val Loss: 0.2476 Acc: 0.9150

Epoch 21/24
----------
train Loss: 0.2881 Acc: 0.8730
val Loss: 0.2396 Acc: 0.9281

Epoch 22/24
----------
train Loss: 0.2263 Acc: 0.9385
val Loss: 0.2348 Acc: 0.9216

Epoch 23/24
----------
train Loss: 0.2500 Acc: 0.8852
val Loss: 0.2533 Acc: 0.9281

Epoch 24/24
----------
train Loss: 0.2726 Acc: 0.8934
val Loss: 0.2399 Acc: 0.9281

Training complete in 1m 10s
Best val Acc: 0.934641

visualize_model(model_ft)

ConvNet as fixed feature extractor

Here, we need to freeze all the network except the final layer. We need to set requires_grad == False to freeze the parameters so that the gradients are not computed in backward().

You can read more about this in the documentation here.

model_conv = torchvision.models.resnet18(pretrained=True)
for param in model_conv.parameters():
    param.requires_grad = False

# Parameters of newly constructed modules have requires_grad=True by default
num_ftrs = model_conv.fc.in_features
model_conv.fc = nn.Linear(num_ftrs, 2)

model_conv = model_conv.to(device)

criterion = nn.CrossEntropyLoss()

# Observe that only parameters of final layer are being optimized as
# opposed to before.
optimizer_conv = optim.SGD(model_conv.fc.parameters(), lr=0.001, momentum=0.9)

# Decay LR by a factor of 0.1 every 7 epochs
exp_lr_scheduler = lr_scheduler.StepLR(optimizer_conv, step_size=7, gamma=0.1)

Train and evaluate

On CPU this will take about half the time compared to previous scenario. This is expected as gradients don’t need to be computed for most of the network. However, forward does need to be computed.

model_conv = train_model(model_conv, criterion, optimizer_conv,
                         exp_lr_scheduler, num_epochs=25)

Out:

Epoch 0/24
----------
train Loss: 0.5475 Acc: 0.7049
val Loss: 0.2228 Acc: 0.9346

Epoch 1/24
----------
train Loss: 0.5115 Acc: 0.7500
val Loss: 0.1902 Acc: 0.9477

Epoch 2/24
----------
train Loss: 0.6709 Acc: 0.7172
val Loss: 0.1790 Acc: 0.9412

Epoch 3/24
----------
train Loss: 0.6136 Acc: 0.7582
val Loss: 0.2086 Acc: 0.9412

Epoch 4/24
----------
train Loss: 0.5677 Acc: 0.7541
val Loss: 0.1837 Acc: 0.9477

Epoch 5/24
----------
train Loss: 0.5572 Acc: 0.7828
val Loss: 0.2249 Acc: 0.9281

Epoch 6/24
----------
train Loss: 0.6167 Acc: 0.7582
val Loss: 0.1822 Acc: 0.9477

Epoch 7/24
----------
train Loss: 0.4875 Acc: 0.7705
val Loss: 0.1784 Acc: 0.9477

Epoch 8/24
----------
train Loss: 0.4510 Acc: 0.7787
val Loss: 0.1803 Acc: 0.9542

Epoch 9/24
----------
train Loss: 0.4077 Acc: 0.8115
val Loss: 0.2130 Acc: 0.9542

Epoch 10/24
----------
train Loss: 0.4449 Acc: 0.8156
val Loss: 0.1671 Acc: 0.9477

Epoch 11/24
----------
train Loss: 0.3937 Acc: 0.8361
val Loss: 0.1738 Acc: 0.9542

Epoch 12/24
----------
train Loss: 0.3437 Acc: 0.8279
val Loss: 0.1774 Acc: 0.9477

Epoch 13/24
----------
train Loss: 0.3363 Acc: 0.8770
val Loss: 0.1814 Acc: 0.9477

Epoch 14/24
----------
train Loss: 0.3235 Acc: 0.8607
val Loss: 0.1764 Acc: 0.9542

Epoch 15/24
----------
train Loss: 0.3990 Acc: 0.8156
val Loss: 0.1865 Acc: 0.9542

Epoch 16/24
----------
train Loss: 0.3650 Acc: 0.8279
val Loss: 0.1655 Acc: 0.9542

Epoch 17/24
----------
train Loss: 0.4488 Acc: 0.8074
val Loss: 0.1767 Acc: 0.9542

Epoch 18/24
----------
train Loss: 0.4024 Acc: 0.8115
val Loss: 0.1831 Acc: 0.9542

Epoch 19/24
----------
train Loss: 0.3496 Acc: 0.8238
val Loss: 0.1875 Acc: 0.9477

Epoch 20/24
----------
train Loss: 0.3476 Acc: 0.8566
val Loss: 0.1993 Acc: 0.9542

Epoch 21/24
----------
train Loss: 0.3115 Acc: 0.8566
val Loss: 0.1907 Acc: 0.9477

Epoch 22/24
----------
train Loss: 0.3791 Acc: 0.8115
val Loss: 0.1726 Acc: 0.9477

Epoch 23/24
----------
train Loss: 0.4003 Acc: 0.8074
val Loss: 0.2054 Acc: 0.9542

Epoch 24/24
----------
train Loss: 0.3599 Acc: 0.8279
val Loss: 0.1880 Acc: 0.9542

Training complete in 0m 33s
Best val Acc: 0.954248

visualize_model(model_conv)

plt.ioff()
plt.show()

Further Learning

If you would like to learn more about the applications of transfer learning, checkout our Quantized Transfer Learning for Computer Vision Tutorial.