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 0x7f528b682860>
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', 'ants', 'bees', 'ants']](../_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]
65%|██████▍ | 28.9M/44.7M [00:00<00:00, 302MB/s]
100%|██████████| 44.7M/44.7M [00:00<00:00, 317MB/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: 0.6206 Acc: 0.6844
val Loss: 0.1918 Acc: 0.9216
Epoch 1/24
----------
train Loss: 0.4515 Acc: 0.8074
val Loss: 0.4496 Acc: 0.8366
Epoch 2/24
----------
train Loss: 0.6664 Acc: 0.7213
val Loss: 0.2994 Acc: 0.8954
Epoch 3/24
----------
train Loss: 0.4359 Acc: 0.7951
val Loss: 0.4481 Acc: 0.8431
Epoch 4/24
----------
train Loss: 0.5663 Acc: 0.8074
val Loss: 0.2550 Acc: 0.9346
Epoch 5/24
----------
train Loss: 0.3821 Acc: 0.8402
val Loss: 0.2908 Acc: 0.9150
Epoch 6/24
----------
train Loss: 0.4602 Acc: 0.8238
val Loss: 0.2971 Acc: 0.8889
Epoch 7/24
----------
train Loss: 0.4049 Acc: 0.8320
val Loss: 0.2252 Acc: 0.9216
Epoch 8/24
----------
train Loss: 0.2932 Acc: 0.8689
val Loss: 0.2598 Acc: 0.9216
Epoch 9/24
----------
train Loss: 0.2358 Acc: 0.8975
val Loss: 0.2369 Acc: 0.9216
Epoch 10/24
----------
train Loss: 0.2774 Acc: 0.8689
val Loss: 0.2303 Acc: 0.9216
Epoch 11/24
----------
train Loss: 0.2726 Acc: 0.8811
val Loss: 0.2241 Acc: 0.9150
Epoch 12/24
----------
train Loss: 0.3135 Acc: 0.8648
val Loss: 0.2189 Acc: 0.9346
Epoch 13/24
----------
train Loss: 0.3472 Acc: 0.8811
val Loss: 0.2262 Acc: 0.9216
Epoch 14/24
----------
train Loss: 0.2678 Acc: 0.8893
val Loss: 0.2239 Acc: 0.9216
Epoch 15/24
----------
train Loss: 0.3281 Acc: 0.8320
val Loss: 0.2248 Acc: 0.9281
Epoch 16/24
----------
train Loss: 0.2837 Acc: 0.8893
val Loss: 0.2177 Acc: 0.9281
Epoch 17/24
----------
train Loss: 0.2592 Acc: 0.8934
val Loss: 0.2200 Acc: 0.9216
Epoch 18/24
----------
train Loss: 0.2789 Acc: 0.8811
val Loss: 0.2208 Acc: 0.9281
Epoch 19/24
----------
train Loss: 0.2683 Acc: 0.8770
val Loss: 0.2225 Acc: 0.9216
Epoch 20/24
----------
train Loss: 0.2348 Acc: 0.9057
val Loss: 0.2243 Acc: 0.9281
Epoch 21/24
----------
train Loss: 0.2696 Acc: 0.9016
val Loss: 0.2174 Acc: 0.9281
Epoch 22/24
----------
train Loss: 0.2279 Acc: 0.9139
val Loss: 0.2316 Acc: 0.9216
Epoch 23/24
----------
train Loss: 0.3294 Acc: 0.8566
val Loss: 0.2174 Acc: 0.9281
Epoch 24/24
----------
train Loss: 0.2729 Acc: 0.8811
val Loss: 0.2244 Acc: 0.9216
Training complete in 0m 34s
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(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: 0.6163 Acc: 0.6393
val Loss: 0.2733 Acc: 0.8627
Epoch 1/24
----------
train Loss: 0.4902 Acc: 0.7705
val Loss: 0.1814 Acc: 0.9412
Epoch 2/24
----------
train Loss: 0.5291 Acc: 0.7418
val Loss: 0.2671 Acc: 0.8954
Epoch 3/24
----------
train Loss: 0.4919 Acc: 0.8033
val Loss: 0.2460 Acc: 0.9085
Epoch 4/24
----------
train Loss: 0.7168 Acc: 0.7254
val Loss: 0.4486 Acc: 0.8366
Epoch 5/24
----------
train Loss: 0.3725 Acc: 0.8320
val Loss: 0.1654 Acc: 0.9542
Epoch 6/24
----------
train Loss: 0.3841 Acc: 0.8074
val Loss: 0.4638 Acc: 0.8366
Epoch 7/24
----------
train Loss: 0.4462 Acc: 0.8320
val Loss: 0.1776 Acc: 0.9412
Epoch 8/24
----------
train Loss: 0.3236 Acc: 0.8607
val Loss: 0.2701 Acc: 0.9020
Epoch 9/24
----------
train Loss: 0.4556 Acc: 0.7664
val Loss: 0.1902 Acc: 0.9412
Epoch 10/24
----------
train Loss: 0.3667 Acc: 0.8607
val Loss: 0.2284 Acc: 0.9216
Epoch 11/24
----------
train Loss: 0.2739 Acc: 0.8770
val Loss: 0.1894 Acc: 0.9477
Epoch 12/24
----------
train Loss: 0.3443 Acc: 0.8566
val Loss: 0.2020 Acc: 0.9346
Epoch 13/24
----------
train Loss: 0.3186 Acc: 0.8566
val Loss: 0.1819 Acc: 0.9412
Epoch 14/24
----------
train Loss: 0.3699 Acc: 0.8484
val Loss: 0.1875 Acc: 0.9412
Epoch 15/24
----------
train Loss: 0.3400 Acc: 0.8484
val Loss: 0.2105 Acc: 0.9346
Epoch 16/24
----------
train Loss: 0.3276 Acc: 0.8402
val Loss: 0.1867 Acc: 0.9412
Epoch 17/24
----------
train Loss: 0.3240 Acc: 0.8770
val Loss: 0.1770 Acc: 0.9542
Epoch 18/24
----------
train Loss: 0.3254 Acc: 0.8525
val Loss: 0.2257 Acc: 0.9281
Epoch 19/24
----------
train Loss: 0.2971 Acc: 0.8730
val Loss: 0.2079 Acc: 0.9412
Epoch 20/24
----------
train Loss: 0.2900 Acc: 0.8852
val Loss: 0.2175 Acc: 0.9216
Epoch 21/24
----------
train Loss: 0.2851 Acc: 0.8443
val Loss: 0.2132 Acc: 0.9281
Epoch 22/24
----------
train Loss: 0.3541 Acc: 0.8320
val Loss: 0.2141 Acc: 0.9216
Epoch 23/24
----------
train Loss: 0.2716 Acc: 0.8852
val Loss: 0.1982 Acc: 0.9346
Epoch 24/24
----------
train Loss: 0.3487 Acc: 0.8484
val Loss: 0.2011 Acc: 0.9346
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 4.437 seconds)
