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 0x7f64b05c3370>
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', '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]
91%|█████████ | 40.5M/44.7M [00:00<00:00, 425MB/s]
100%|██████████| 44.7M/44.7M [00:00<00:00, 426MB/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: 0.6229 Acc: 0.7213
val Loss: 0.2318 Acc: 0.9216
Epoch 1/24
----------
train Loss: 0.6955 Acc: 0.6885
val Loss: 0.2315 Acc: 0.9085
Epoch 2/24
----------
train Loss: 0.6665 Acc: 0.7418
val Loss: 0.2108 Acc: 0.9346
Epoch 3/24
----------
train Loss: 0.5244 Acc: 0.8115
val Loss: 0.5298 Acc: 0.8366
Epoch 4/24
----------
train Loss: 0.5970 Acc: 0.7582
val Loss: 0.5967 Acc: 0.7451
Epoch 5/24
----------
train Loss: 0.4371 Acc: 0.8361
val Loss: 0.2922 Acc: 0.8693
Epoch 6/24
----------
train Loss: 0.5674 Acc: 0.7664
val Loss: 0.3279 Acc: 0.8627
Epoch 7/24
----------
train Loss: 0.3649 Acc: 0.8566
val Loss: 0.2387 Acc: 0.9150
Epoch 8/24
----------
train Loss: 0.3884 Acc: 0.8484
val Loss: 0.2225 Acc: 0.9150
Epoch 9/24
----------
train Loss: 0.4063 Acc: 0.7992
val Loss: 0.3170 Acc: 0.8497
Epoch 10/24
----------
train Loss: 0.2582 Acc: 0.8730
val Loss: 0.2128 Acc: 0.9216
Epoch 11/24
----------
train Loss: 0.3158 Acc: 0.8730
val Loss: 0.2212 Acc: 0.9085
Epoch 12/24
----------
train Loss: 0.3275 Acc: 0.8443
val Loss: 0.2107 Acc: 0.9150
Epoch 13/24
----------
train Loss: 0.2847 Acc: 0.8730
val Loss: 0.2207 Acc: 0.9020
Epoch 14/24
----------
train Loss: 0.1782 Acc: 0.9426
val Loss: 0.1996 Acc: 0.9216
Epoch 15/24
----------
train Loss: 0.2223 Acc: 0.8934
val Loss: 0.1997 Acc: 0.9150
Epoch 16/24
----------
train Loss: 0.2765 Acc: 0.8730
val Loss: 0.2010 Acc: 0.9085
Epoch 17/24
----------
train Loss: 0.2809 Acc: 0.8607
val Loss: 0.2146 Acc: 0.9150
Epoch 18/24
----------
train Loss: 0.3457 Acc: 0.8607
val Loss: 0.2066 Acc: 0.9150
Epoch 19/24
----------
train Loss: 0.3623 Acc: 0.8402
val Loss: 0.2025 Acc: 0.9150
Epoch 20/24
----------
train Loss: 0.3433 Acc: 0.8607
val Loss: 0.2130 Acc: 0.9020
Epoch 21/24
----------
train Loss: 0.2213 Acc: 0.9221
val Loss: 0.2254 Acc: 0.9216
Epoch 22/24
----------
train Loss: 0.2951 Acc: 0.8811
val Loss: 0.2012 Acc: 0.9216
Epoch 23/24
----------
train Loss: 0.2640 Acc: 0.8893
val Loss: 0.2195 Acc: 0.9085
Epoch 24/24
----------
train Loss: 0.2251 Acc: 0.9057
val Loss: 0.2242 Acc: 0.9150
Training complete in 0m 36s
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: 0.5256 Acc: 0.7254
val Loss: 0.2735 Acc: 0.8954
Epoch 1/24
----------
train Loss: 0.5033 Acc: 0.7828
val Loss: 1.3175 Acc: 0.5621
Epoch 2/24
----------
train Loss: 0.7428 Acc: 0.7705
val Loss: 0.2103 Acc: 0.9281
Epoch 3/24
----------
train Loss: 0.5383 Acc: 0.7705
val Loss: 0.5245 Acc: 0.8105
Epoch 4/24
----------
train Loss: 0.5406 Acc: 0.7787
val Loss: 0.2201 Acc: 0.9412
Epoch 5/24
----------
train Loss: 0.4546 Acc: 0.8279
val Loss: 0.1917 Acc: 0.9542
Epoch 6/24
----------
train Loss: 0.4908 Acc: 0.8074
val Loss: 0.1968 Acc: 0.9346
Epoch 7/24
----------
train Loss: 0.3160 Acc: 0.8811
val Loss: 0.2180 Acc: 0.9281
Epoch 8/24
----------
train Loss: 0.3436 Acc: 0.8730
val Loss: 0.1945 Acc: 0.9346
Epoch 9/24
----------
train Loss: 0.3876 Acc: 0.8320
val Loss: 0.1944 Acc: 0.9412
Epoch 10/24
----------
train Loss: 0.2757 Acc: 0.8934
val Loss: 0.2416 Acc: 0.9085
Epoch 11/24
----------
train Loss: 0.2679 Acc: 0.8770
val Loss: 0.1863 Acc: 0.9412
Epoch 12/24
----------
train Loss: 0.3448 Acc: 0.8648
val Loss: 0.1871 Acc: 0.9412
Epoch 13/24
----------
train Loss: 0.1942 Acc: 0.9262
val Loss: 0.2183 Acc: 0.9281
Epoch 14/24
----------
train Loss: 0.3933 Acc: 0.8443
val Loss: 0.1943 Acc: 0.9412
Epoch 15/24
----------
train Loss: 0.3667 Acc: 0.8525
val Loss: 0.1995 Acc: 0.9346
Epoch 16/24
----------
train Loss: 0.3705 Acc: 0.8402
val Loss: 0.1785 Acc: 0.9412
Epoch 17/24
----------
train Loss: 0.2630 Acc: 0.8852
val Loss: 0.2075 Acc: 0.9412
Epoch 18/24
----------
train Loss: 0.3376 Acc: 0.8443
val Loss: 0.2918 Acc: 0.8824
Epoch 19/24
----------
train Loss: 0.2968 Acc: 0.8689
val Loss: 0.1928 Acc: 0.9346
Epoch 20/24
----------
train Loss: 0.3954 Acc: 0.7992
val Loss: 0.1755 Acc: 0.9412
Epoch 21/24
----------
train Loss: 0.3202 Acc: 0.8566
val Loss: 0.2193 Acc: 0.9281
Epoch 22/24
----------
train Loss: 0.3457 Acc: 0.8689
val Loss: 0.1867 Acc: 0.9281
Epoch 23/24
----------
train Loss: 0.3767 Acc: 0.8320
val Loss: 0.1817 Acc: 0.9412
Epoch 24/24
----------
train Loss: 0.4384 Acc: 0.8320
val Loss: 0.1964 Acc: 0.9281
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 6.620 seconds)