beginner/transfer_learning_tutorial

Run in Google Colab

Colab

Download Notebook

Notebook

View on GitHub

GitHub

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 0x7f50c7d78b80> 

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]) 

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] 88%|████████▊ | 39.1M/44.7M [00:00<00:00, 410MB/s] 100%|██████████| 44.7M/44.7M [00:00<00:00, 414MB/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: 0.6921 Acc: 0.6967 val Loss: 0.8499 Acc: 0.6993 Epoch 1/24 ---------- train Loss: 0.5841 Acc: 0.7664 val Loss: 0.3710 Acc: 0.8889 Epoch 2/24 ---------- train Loss: 0.5751 Acc: 0.7500 val Loss: 0.5601 Acc: 0.8235 Epoch 3/24 ---------- train Loss: 0.7188 Acc: 0.7254 val Loss: 0.4344 Acc: 0.8824 Epoch 4/24 ---------- train Loss: 0.6559 Acc: 0.7500 val Loss: 0.5307 Acc: 0.8758 Epoch 5/24 ---------- train Loss: 0.5357 Acc: 0.8115 val Loss: 0.4136 Acc: 0.8889 Epoch 6/24 ---------- train Loss: 0.5735 Acc: 0.8156 val Loss: 0.6544 Acc: 0.8235 Epoch 7/24 ---------- train Loss: 0.3559 Acc: 0.8689 val Loss: 0.3867 Acc: 0.8758 Epoch 8/24 ---------- train Loss: 0.2945 Acc: 0.8893 val Loss: 0.3452 Acc: 0.8954 Epoch 9/24 ---------- train Loss: 0.3971 Acc: 0.8607 val Loss: 0.3640 Acc: 0.8954 Epoch 10/24 ---------- train Loss: 0.3288 Acc: 0.8811 val Loss: 0.3327 Acc: 0.8824 Epoch 11/24 ---------- train Loss: 0.3723 Acc: 0.8402 val Loss: 0.3053 Acc: 0.8889 Epoch 12/24 ---------- train Loss: 0.3929 Acc: 0.8566 val Loss: 0.3225 Acc: 0.9085 Epoch 13/24 ---------- train Loss: 0.3996 Acc: 0.8443 val Loss: 0.2910 Acc: 0.9085 Epoch 14/24 ---------- train Loss: 0.2153 Acc: 0.9098 val Loss: 0.2990 Acc: 0.9020 Epoch 15/24 ---------- train Loss: 0.2831 Acc: 0.8648 val Loss: 0.2949 Acc: 0.9020 Epoch 16/24 ---------- train Loss: 0.1627 Acc: 0.9344 val Loss: 0.2983 Acc: 0.9085 Epoch 17/24 ---------- train Loss: 0.3799 Acc: 0.8525 val Loss: 0.3126 Acc: 0.9020 Epoch 18/24 ---------- train Loss: 0.2147 Acc: 0.9180 val Loss: 0.3134 Acc: 0.9216 Epoch 19/24 ---------- train Loss: 0.2689 Acc: 0.8852 val Loss: 0.2818 Acc: 0.9150 Epoch 20/24 ---------- train Loss: 0.1948 Acc: 0.9262 val Loss: 0.2943 Acc: 0.8889 Epoch 21/24 ---------- train Loss: 0.2448 Acc: 0.9221 val Loss: 0.2948 Acc: 0.9085 Epoch 22/24 ---------- train Loss: 0.2571 Acc: 0.8934 val Loss: 0.3128 Acc: 0.8824 Epoch 23/24 ---------- train Loss: 0.2568 Acc: 0.8730 val Loss: 0.2890 Acc: 0.9150 Epoch 24/24 ---------- train Loss: 0.2021 Acc: 0.8893 val Loss: 0.2846 Acc: 0.9150 Training complete in 0m 35s 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: 0.6217 Acc: 0.6598 val Loss: 0.2205 Acc: 0.9542 Epoch 1/24 ---------- train Loss: 0.4535 Acc: 0.7992 val Loss: 0.1925 Acc: 0.9412 Epoch 2/24 ---------- train Loss: 0.4486 Acc: 0.7910 val Loss: 0.2046 Acc: 0.9216 Epoch 3/24 ---------- train Loss: 0.5006 Acc: 0.7705 val Loss: 0.1846 Acc: 0.9542 Epoch 4/24 ---------- train Loss: 0.4839 Acc: 0.7951 val Loss: 1.2583 Acc: 0.6275 Epoch 5/24 ---------- train Loss: 0.7651 Acc: 0.7746 val Loss: 0.2445 Acc: 0.9150 Epoch 6/24 ---------- train Loss: 0.4886 Acc: 0.7787 val Loss: 0.2071 Acc: 0.9542 Epoch 7/24 ---------- train Loss: 0.3475 Acc: 0.8730 val Loss: 0.1708 Acc: 0.9477 Epoch 8/24 ---------- train Loss: 0.3603 Acc: 0.8443 val Loss: 0.2032 Acc: 0.9412 Epoch 9/24 ---------- train Loss: 0.3865 Acc: 0.8279 val Loss: 0.1809 Acc: 0.9477 Epoch 10/24 ---------- train Loss: 0.3242 Acc: 0.8852 val Loss: 0.1839 Acc: 0.9477 Epoch 11/24 ---------- train Loss: 0.4182 Acc: 0.8197 val Loss: 0.1711 Acc: 0.9608 Epoch 12/24 ---------- train Loss: 0.2985 Acc: 0.8770 val Loss: 0.2023 Acc: 0.9412 Epoch 13/24 ---------- train Loss: 0.3571 Acc: 0.8238 val Loss: 0.1754 Acc: 0.9477 Epoch 14/24 ---------- train Loss: 0.3897 Acc: 0.8484 val Loss: 0.1809 Acc: 0.9412 Epoch 15/24 ---------- train Loss: 0.3077 Acc: 0.8607 val Loss: 0.1807 Acc: 0.9477 Epoch 16/24 ---------- train Loss: 0.3589 Acc: 0.8484 val Loss: 0.1745 Acc: 0.9412 Epoch 17/24 ---------- train Loss: 0.2535 Acc: 0.8730 val Loss: 0.1638 Acc: 0.9477 Epoch 18/24 ---------- train Loss: 0.3539 Acc: 0.8320 val Loss: 0.1757 Acc: 0.9477 Epoch 19/24 ---------- train Loss: 0.3944 Acc: 0.8156 val Loss: 0.1706 Acc: 0.9542 Epoch 20/24 ---------- train Loss: 0.3395 Acc: 0.8648 val Loss: 0.2025 Acc: 0.9412 Epoch 21/24 ---------- train Loss: 0.3783 Acc: 0.8443 val Loss: 0.1748 Acc: 0.9412 Epoch 22/24 ---------- train Loss: 0.4309 Acc: 0.8156 val Loss: 0.1824 Acc: 0.9412 Epoch 23/24 ---------- train Loss: 0.3715 Acc: 0.8279 val Loss: 0.1837 Acc: 0.9542 Epoch 24/24 ---------- train Loss: 0.3148 Acc: 0.8443 val Loss: 0.1971 Acc: 0.9542 Training complete in 0m 27s 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.