Note
Go to the end to download the full example code.
Transfer Learning for Computer Vision Tutorial#
Created On: Mar 24, 2017 | Last Updated: Jan 27, 2025 | Last Verified: Nov 05, 2024
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 initialization, 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
import torch
import torch.nn as nn
import torch.optim as optim
from torch.optim import lr_scheduler
import torch.backends.cudnn as cudnn
import numpy as np
import torchvision
from torchvision import datasets, models, transforms
import matplotlib.pyplot as plt
import time
import os
from PIL import Image
from tempfile import TemporaryDirectory
cudnn.benchmark = True
plt.ion() # interactive mode
<contextlib.ExitStack object at 0x7fb95fd58970>
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
# We want to be able to train our model on an `accelerator <https://pytorch.org/docs/stable/torch.html#accelerators>`__
# such as CUDA, MPS, MTIA, or XPU. If the current accelerator is available, we will use it. Otherwise, we use the CPU.
device = torch.accelerator.current_accelerator().type if torch.accelerator.is_available() else "cpu"
print(f"Using {device} device")
Using cuda device
Visualize a few images#
Let’s visualize a few training images so as to understand the data augmentations.
def imshow(inp, title=None):
"""Display image 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])
![['ants', 'bees', 'bees', 'bees']](../_images/sphx_glr_transfer_learning_tutorial_001.png)
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()
# Create a temporary directory to save training checkpoints
with TemporaryDirectory() as tempdir:
best_model_params_path = os.path.join(tempdir, 'best_model_params.pt')
torch.save(model.state_dict(), best_model_params_path)
best_acc = 0.0
for epoch in range(num_epochs):
print(f'Epoch {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(f'{phase} Loss: {epoch_loss:.4f} Acc: {epoch_acc:.4f}')
# deep copy the model
if phase == 'val' and epoch_acc > best_acc:
best_acc = epoch_acc
torch.save(model.state_dict(), best_model_params_path)
print()
time_elapsed = time.time() - since
print(f'Training complete in {time_elapsed // 60:.0f}m {time_elapsed % 60:.0f}s')
print(f'Best val Acc: {best_acc:4f}')
# load best model weights
model.load_state_dict(torch.load(best_model_params_path, weights_only=True))
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(f'predicted: {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(weights='IMAGENET1K_V1')
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)
Downloading: "https://download.pytorch.org/models/resnet18-f37072fd.pth" to /var/lib/ci-user/.cache/torch/hub/checkpoints/resnet18-f37072fd.pth
0%| | 0.00/44.7M [00:00<?, ?B/s]
83%|████████▎ | 37.2M/44.7M [00:00<00:00, 390MB/s]
100%|██████████| 44.7M/44.7M [00:00<00:00, 395MB/s]
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)
Epoch 0/24
----------
train Loss: 0.5961 Acc: 0.7131
val Loss: 0.2213 Acc: 0.9085
Epoch 1/24
----------
train Loss: 0.5420 Acc: 0.8115
val Loss: 0.4113 Acc: 0.8301
Epoch 2/24
----------
train Loss: 0.4799 Acc: 0.7910
val Loss: 0.3714 Acc: 0.8431
Epoch 3/24
----------
train Loss: 0.3895 Acc: 0.8607
val Loss: 0.2667 Acc: 0.9020
Epoch 4/24
----------
train Loss: 0.3358 Acc: 0.8566
val Loss: 0.3453 Acc: 0.9020
Epoch 5/24
----------
train Loss: 0.4386 Acc: 0.8279
val Loss: 0.8397 Acc: 0.7843
Epoch 6/24
----------
train Loss: 0.6263 Acc: 0.7664
val Loss: 0.3024 Acc: 0.8889
Epoch 7/24
----------
train Loss: 0.3348 Acc: 0.8607
val Loss: 0.2581 Acc: 0.9085
Epoch 8/24
----------
train Loss: 0.3651 Acc: 0.8361
val Loss: 0.2399 Acc: 0.9216
Epoch 9/24
----------
train Loss: 0.3738 Acc: 0.8484
val Loss: 0.2640 Acc: 0.9020
Epoch 10/24
----------
train Loss: 0.3494 Acc: 0.8689
val Loss: 0.2348 Acc: 0.8954
Epoch 11/24
----------
train Loss: 0.3192 Acc: 0.8525
val Loss: 0.2324 Acc: 0.9085
Epoch 12/24
----------
train Loss: 0.2844 Acc: 0.8770
val Loss: 0.2307 Acc: 0.9085
Epoch 13/24
----------
train Loss: 0.2395 Acc: 0.8975
val Loss: 0.2053 Acc: 0.9412
Epoch 14/24
----------
train Loss: 0.2914 Acc: 0.8607
val Loss: 0.2205 Acc: 0.9346
Epoch 15/24
----------
train Loss: 0.2536 Acc: 0.8934
val Loss: 0.2249 Acc: 0.9085
Epoch 16/24
----------
train Loss: 0.2613 Acc: 0.8975
val Loss: 0.2255 Acc: 0.9150
Epoch 17/24
----------
train Loss: 0.2743 Acc: 0.8893
val Loss: 0.2222 Acc: 0.9216
Epoch 18/24
----------
train Loss: 0.3101 Acc: 0.8484
val Loss: 0.2136 Acc: 0.9216
Epoch 19/24
----------
train Loss: 0.2538 Acc: 0.9057
val Loss: 0.2100 Acc: 0.9216
Epoch 20/24
----------
train Loss: 0.2950 Acc: 0.8689
val Loss: 0.2291 Acc: 0.9020
Epoch 21/24
----------
train Loss: 0.2959 Acc: 0.8730
val Loss: 0.2242 Acc: 0.9150
Epoch 22/24
----------
train Loss: 0.2467 Acc: 0.9016
val Loss: 0.2189 Acc: 0.9346
Epoch 23/24
----------
train Loss: 0.2570 Acc: 0.8893
val Loss: 0.2031 Acc: 0.9281
Epoch 24/24
----------
train Loss: 0.2158 Acc: 0.9262
val Loss: 0.2161 Acc: 0.9346
Training complete in 0m 36s
Best val Acc: 0.941176
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(weights='IMAGENET1K_V1')
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)
Epoch 0/24
----------
train Loss: 0.7416 Acc: 0.6230
val Loss: 1.0094 Acc: 0.5817
Epoch 1/24
----------
train Loss: 0.5572 Acc: 0.7582
val Loss: 0.2447 Acc: 0.9085
Epoch 2/24
----------
train Loss: 0.5675 Acc: 0.7459
val Loss: 0.2690 Acc: 0.8889
Epoch 3/24
----------
train Loss: 0.5289 Acc: 0.7664
val Loss: 0.1971 Acc: 0.9281
Epoch 4/24
----------
train Loss: 0.4397 Acc: 0.8279
val Loss: 0.1633 Acc: 0.9477
Epoch 5/24
----------
train Loss: 0.4192 Acc: 0.8320
val Loss: 0.2027 Acc: 0.9346
Epoch 6/24
----------
train Loss: 0.3932 Acc: 0.8320
val Loss: 0.2797 Acc: 0.8954
Epoch 7/24
----------
train Loss: 0.2777 Acc: 0.9057
val Loss: 0.1619 Acc: 0.9542
Epoch 8/24
----------
train Loss: 0.2683 Acc: 0.8648
val Loss: 0.1949 Acc: 0.9477
Epoch 9/24
----------
train Loss: 0.2872 Acc: 0.8811
val Loss: 0.1677 Acc: 0.9542
Epoch 10/24
----------
train Loss: 0.3619 Acc: 0.8443
val Loss: 0.2375 Acc: 0.9346
Epoch 11/24
----------
train Loss: 0.4120 Acc: 0.7869
val Loss: 0.2016 Acc: 0.9412
Epoch 12/24
----------
train Loss: 0.3496 Acc: 0.8320
val Loss: 0.1806 Acc: 0.9477
Epoch 13/24
----------
train Loss: 0.3622 Acc: 0.8320
val Loss: 0.1910 Acc: 0.9477
Epoch 14/24
----------
train Loss: 0.2892 Acc: 0.8525
val Loss: 0.1735 Acc: 0.9412
Epoch 15/24
----------
train Loss: 0.3057 Acc: 0.8689
val Loss: 0.2025 Acc: 0.9412
Epoch 16/24
----------
train Loss: 0.3252 Acc: 0.8730
val Loss: 0.1747 Acc: 0.9542
Epoch 17/24
----------
train Loss: 0.3232 Acc: 0.8730
val Loss: 0.1783 Acc: 0.9412
Epoch 18/24
----------
train Loss: 0.3613 Acc: 0.8443
val Loss: 0.1921 Acc: 0.9477
Epoch 19/24
----------
train Loss: 0.2884 Acc: 0.8811
val Loss: 0.1846 Acc: 0.9477
Epoch 20/24
----------
train Loss: 0.4015 Acc: 0.8566
val Loss: 0.1637 Acc: 0.9542
Epoch 21/24
----------
train Loss: 0.2969 Acc: 0.8770
val Loss: 0.1725 Acc: 0.9542
Epoch 22/24
----------
train Loss: 0.2879 Acc: 0.8975
val Loss: 0.1915 Acc: 0.9477
Epoch 23/24
----------
train Loss: 0.3895 Acc: 0.8238
val Loss: 0.1699 Acc: 0.9477
Epoch 24/24
----------
train Loss: 0.3719 Acc: 0.8402
val Loss: 0.1627 Acc: 0.9412
Training complete in 0m 28s
Best val Acc: 0.954248
visualize_model(model_conv)
plt.ioff()
plt.show()
Inference on custom images#
Use the trained model to make predictions on custom images and visualize the predicted class labels along with the images.
def visualize_model_predictions(model,img_path):
was_training = model.training
model.eval()
img = Image.open(img_path)
img = data_transforms['val'](img)
img = img.unsqueeze(0)
img = img.to(device)
with torch.no_grad():
outputs = model(img)
_, preds = torch.max(outputs, 1)
ax = plt.subplot(2,2,1)
ax.axis('off')
ax.set_title(f'Predicted: {class_names[preds[0]]}')
imshow(img.cpu().data[0])
model.train(mode=was_training)
visualize_model_predictions(
model_conv,
img_path='data/hymenoptera_data/val/bees/72100438_73de9f17af.jpg'
)
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.
Total running time of the script: (1 minutes 6.807 seconds)