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 0x7f276a687af0>
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', '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]
90%|████████▉ | 40.0M/44.7M [00:00<00:00, 419MB/s]
100%|██████████| 44.7M/44.7M [00:00<00:00, 421MB/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: 0.7185 Acc: 0.6639
val Loss: 0.4354 Acc: 0.8627
Epoch 1/24
----------
train Loss: 0.4421 Acc: 0.8115
val Loss: 0.4074 Acc: 0.8627
Epoch 2/24
----------
train Loss: 0.6345 Acc: 0.7623
val Loss: 0.3004 Acc: 0.8954
Epoch 3/24
----------
train Loss: 0.6572 Acc: 0.7213
val Loss: 0.3164 Acc: 0.8758
Epoch 4/24
----------
train Loss: 0.3708 Acc: 0.8484
val Loss: 0.4053 Acc: 0.9150
Epoch 5/24
----------
train Loss: 0.6032 Acc: 0.7500
val Loss: 0.5950 Acc: 0.8497
Epoch 6/24
----------
train Loss: 0.4314 Acc: 0.8156
val Loss: 0.3230 Acc: 0.8889
Epoch 7/24
----------
train Loss: 0.4482 Acc: 0.8279
val Loss: 0.3204 Acc: 0.8954
Epoch 8/24
----------
train Loss: 0.3634 Acc: 0.8320
val Loss: 0.2953 Acc: 0.9020
Epoch 9/24
----------
train Loss: 0.4052 Acc: 0.8197
val Loss: 0.2921 Acc: 0.8954
Epoch 10/24
----------
train Loss: 0.2804 Acc: 0.8975
val Loss: 0.3427 Acc: 0.9020
Epoch 11/24
----------
train Loss: 0.3185 Acc: 0.8811
val Loss: 0.2904 Acc: 0.9216
Epoch 12/24
----------
train Loss: 0.3549 Acc: 0.8197
val Loss: 0.2919 Acc: 0.9020
Epoch 13/24
----------
train Loss: 0.2493 Acc: 0.8770
val Loss: 0.2850 Acc: 0.9150
Epoch 14/24
----------
train Loss: 0.2618 Acc: 0.8811
val Loss: 0.2914 Acc: 0.9085
Epoch 15/24
----------
train Loss: 0.2886 Acc: 0.8525
val Loss: 0.3161 Acc: 0.9150
Epoch 16/24
----------
train Loss: 0.3088 Acc: 0.8770
val Loss: 0.2885 Acc: 0.9085
Epoch 17/24
----------
train Loss: 0.3062 Acc: 0.8607
val Loss: 0.2730 Acc: 0.9085
Epoch 18/24
----------
train Loss: 0.2827 Acc: 0.8852
val Loss: 0.2680 Acc: 0.9150
Epoch 19/24
----------
train Loss: 0.3489 Acc: 0.8484
val Loss: 0.2874 Acc: 0.9085
Epoch 20/24
----------
train Loss: 0.3242 Acc: 0.8525
val Loss: 0.2977 Acc: 0.9020
Epoch 21/24
----------
train Loss: 0.2354 Acc: 0.8730
val Loss: 0.2794 Acc: 0.9150
Epoch 22/24
----------
train Loss: 0.3410 Acc: 0.8525
val Loss: 0.2763 Acc: 0.9020
Epoch 23/24
----------
train Loss: 0.2497 Acc: 0.8893
val Loss: 0.2733 Acc: 0.9216
Epoch 24/24
----------
train Loss: 0.3127 Acc: 0.8689
val Loss: 0.2725 Acc: 0.9150
Training complete in 0m 36s
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: 0.5794 Acc: 0.6680
val Loss: 0.2769 Acc: 0.9020
Epoch 1/24
----------
train Loss: 0.5033 Acc: 0.7623
val Loss: 0.2196 Acc: 0.9216
Epoch 2/24
----------
train Loss: 0.4193 Acc: 0.7828
val Loss: 0.2133 Acc: 0.9477
Epoch 3/24
----------
train Loss: 0.4552 Acc: 0.8115
val Loss: 0.2127 Acc: 0.9477
Epoch 4/24
----------
train Loss: 0.3278 Acc: 0.8484
val Loss: 0.2630 Acc: 0.9085
Epoch 5/24
----------
train Loss: 0.5544 Acc: 0.7787
val Loss: 0.1852 Acc: 0.9412
Epoch 6/24
----------
train Loss: 0.3605 Acc: 0.8484
val Loss: 0.2286 Acc: 0.9150
Epoch 7/24
----------
train Loss: 0.5108 Acc: 0.7664
val Loss: 0.2175 Acc: 0.9281
Epoch 8/24
----------
train Loss: 0.3629 Acc: 0.8566
val Loss: 0.1818 Acc: 0.9412
Epoch 9/24
----------
train Loss: 0.3406 Acc: 0.8320
val Loss: 0.1788 Acc: 0.9477
Epoch 10/24
----------
train Loss: 0.3604 Acc: 0.8320
val Loss: 0.1907 Acc: 0.9216
Epoch 11/24
----------
train Loss: 0.2992 Acc: 0.8648
val Loss: 0.1743 Acc: 0.9412
Epoch 12/24
----------
train Loss: 0.3982 Acc: 0.8361
val Loss: 0.1960 Acc: 0.9281
Epoch 13/24
----------
train Loss: 0.2968 Acc: 0.8730
val Loss: 0.1880 Acc: 0.9477
Epoch 14/24
----------
train Loss: 0.3424 Acc: 0.8525
val Loss: 0.1854 Acc: 0.9608
Epoch 15/24
----------
train Loss: 0.2849 Acc: 0.8730
val Loss: 0.1685 Acc: 0.9346
Epoch 16/24
----------
train Loss: 0.3713 Acc: 0.8484
val Loss: 0.1863 Acc: 0.9477
Epoch 17/24
----------
train Loss: 0.3511 Acc: 0.8525
val Loss: 0.1933 Acc: 0.9346
Epoch 18/24
----------
train Loss: 0.3785 Acc: 0.8402
val Loss: 0.1880 Acc: 0.9412
Epoch 19/24
----------
train Loss: 0.3149 Acc: 0.8811
val Loss: 0.1836 Acc: 0.9412
Epoch 20/24
----------
train Loss: 0.3307 Acc: 0.8443
val Loss: 0.1832 Acc: 0.9477
Epoch 21/24
----------
train Loss: 0.3149 Acc: 0.8607
val Loss: 0.1776 Acc: 0.9412
Epoch 22/24
----------
train Loss: 0.3054 Acc: 0.8730
val Loss: 0.1848 Acc: 0.9412
Epoch 23/24
----------
train Loss: 0.2893 Acc: 0.8730
val Loss: 0.1691 Acc: 0.9477
Epoch 24/24
----------
train Loss: 0.3339 Acc: 0.8484
val Loss: 0.1752 Acc: 0.9412
Training complete in 0m 28s
Best val Acc: 0.960784
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 5.615 seconds)