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 0x7fd3cc804490>
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])
![['bees', 'bees', 'ants', '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]
67%|██████▋ | 29.9M/44.7M [00:00<00:00, 313MB/s]
100%|██████████| 44.7M/44.7M [00:00<00:00, 325MB/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: 0.5795 Acc: 0.6844
val Loss: 0.2221 Acc: 0.9216
Epoch 1/24
----------
train Loss: 0.3820 Acc: 0.8525
val Loss: 0.3049 Acc: 0.9150
Epoch 2/24
----------
train Loss: 0.6699 Acc: 0.7910
val Loss: 0.5974 Acc: 0.8758
Epoch 3/24
----------
train Loss: 0.5291 Acc: 0.7951
val Loss: 0.2780 Acc: 0.8889
Epoch 4/24
----------
train Loss: 0.4411 Acc: 0.7951
val Loss: 0.6140 Acc: 0.8105
Epoch 5/24
----------
train Loss: 0.5920 Acc: 0.8074
val Loss: 0.2255 Acc: 0.9085
Epoch 6/24
----------
train Loss: 0.5488 Acc: 0.8074
val Loss: 0.3690 Acc: 0.8889
Epoch 7/24
----------
train Loss: 0.3998 Acc: 0.8607
val Loss: 0.2724 Acc: 0.9020
Epoch 8/24
----------
train Loss: 0.3069 Acc: 0.8484
val Loss: 0.2734 Acc: 0.8954
Epoch 9/24
----------
train Loss: 0.2630 Acc: 0.8770
val Loss: 0.2279 Acc: 0.9085
Epoch 10/24
----------
train Loss: 0.2158 Acc: 0.9180
val Loss: 0.2345 Acc: 0.9085
Epoch 11/24
----------
train Loss: 0.2390 Acc: 0.8975
val Loss: 0.2269 Acc: 0.9085
Epoch 12/24
----------
train Loss: 0.3001 Acc: 0.8934
val Loss: 0.2194 Acc: 0.9085
Epoch 13/24
----------
train Loss: 0.2542 Acc: 0.8852
val Loss: 0.2094 Acc: 0.9085
Epoch 14/24
----------
train Loss: 0.2115 Acc: 0.9016
val Loss: 0.2239 Acc: 0.9020
Epoch 15/24
----------
train Loss: 0.2815 Acc: 0.8770
val Loss: 0.2500 Acc: 0.9150
Epoch 16/24
----------
train Loss: 0.2352 Acc: 0.9098
val Loss: 0.2418 Acc: 0.8954
Epoch 17/24
----------
train Loss: 0.2684 Acc: 0.8770
val Loss: 0.2332 Acc: 0.9216
Epoch 18/24
----------
train Loss: 0.2783 Acc: 0.8893
val Loss: 0.2408 Acc: 0.9085
Epoch 19/24
----------
train Loss: 0.2440 Acc: 0.8934
val Loss: 0.2219 Acc: 0.9150
Epoch 20/24
----------
train Loss: 0.1853 Acc: 0.9262
val Loss: 0.2146 Acc: 0.9150
Epoch 21/24
----------
train Loss: 0.2058 Acc: 0.9016
val Loss: 0.2225 Acc: 0.9085
Epoch 22/24
----------
train Loss: 0.3241 Acc: 0.8730
val Loss: 0.2173 Acc: 0.9216
Epoch 23/24
----------
train Loss: 0.2111 Acc: 0.9221
val Loss: 0.2369 Acc: 0.9150
Epoch 24/24
----------
train Loss: 0.3334 Acc: 0.8689
val Loss: 0.2154 Acc: 0.9020
Training complete in 0m 34s
Best val Acc: 0.921569
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: 0.5421 Acc: 0.6885
val Loss: 0.2343 Acc: 0.9477
Epoch 1/24
----------
train Loss: 0.4008 Acc: 0.8238
val Loss: 0.2424 Acc: 0.9216
Epoch 2/24
----------
train Loss: 0.4481 Acc: 0.7664
val Loss: 0.1925 Acc: 0.9346
Epoch 3/24
----------
train Loss: 0.4680 Acc: 0.7541
val Loss: 0.2885 Acc: 0.8758
Epoch 4/24
----------
train Loss: 0.4371 Acc: 0.8197
val Loss: 0.3388 Acc: 0.8562
Epoch 5/24
----------
train Loss: 0.3789 Acc: 0.8402
val Loss: 0.2743 Acc: 0.8889
Epoch 6/24
----------
train Loss: 0.3840 Acc: 0.8525
val Loss: 0.2769 Acc: 0.8758
Epoch 7/24
----------
train Loss: 0.4059 Acc: 0.8238
val Loss: 0.1974 Acc: 0.9281
Epoch 8/24
----------
train Loss: 0.3928 Acc: 0.8238
val Loss: 0.1898 Acc: 0.9477
Epoch 9/24
----------
train Loss: 0.3436 Acc: 0.8525
val Loss: 0.1867 Acc: 0.9412
Epoch 10/24
----------
train Loss: 0.3974 Acc: 0.8238
val Loss: 0.1793 Acc: 0.9542
Epoch 11/24
----------
train Loss: 0.3407 Acc: 0.8525
val Loss: 0.1680 Acc: 0.9542
Epoch 12/24
----------
train Loss: 0.3999 Acc: 0.8115
val Loss: 0.1637 Acc: 0.9412
Epoch 13/24
----------
train Loss: 0.3803 Acc: 0.8033
val Loss: 0.1808 Acc: 0.9477
Epoch 14/24
----------
train Loss: 0.3742 Acc: 0.8443
val Loss: 0.1909 Acc: 0.9542
Epoch 15/24
----------
train Loss: 0.3334 Acc: 0.8525
val Loss: 0.1775 Acc: 0.9542
Epoch 16/24
----------
train Loss: 0.3342 Acc: 0.8607
val Loss: 0.2232 Acc: 0.9150
Epoch 17/24
----------
train Loss: 0.2972 Acc: 0.8934
val Loss: 0.1874 Acc: 0.9477
Epoch 18/24
----------
train Loss: 0.3442 Acc: 0.8648
val Loss: 0.1743 Acc: 0.9542
Epoch 19/24
----------
train Loss: 0.3373 Acc: 0.8443
val Loss: 0.1802 Acc: 0.9542
Epoch 20/24
----------
train Loss: 0.3413 Acc: 0.8525
val Loss: 0.2075 Acc: 0.9216
Epoch 21/24
----------
train Loss: 0.2268 Acc: 0.9262
val Loss: 0.2246 Acc: 0.9020
Epoch 22/24
----------
train Loss: 0.3271 Acc: 0.8238
val Loss: 0.1689 Acc: 0.9542
Epoch 23/24
----------
train Loss: 0.3793 Acc: 0.8320
val Loss: 0.1796 Acc: 0.9477
Epoch 24/24
----------
train Loss: 0.3014 Acc: 0.8607
val Loss: 0.1882 Acc: 0.9477
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 5.096 seconds)
