迁移学习在图像分类中的应用

常言道“熟能生巧”,通过大量练习不同领域的题目来为最终的大考做准备。如果能将这种技巧应用于分类、回归或聚类问题,那会怎样呢?迁移学习就是这样一种技术,它允许利用在标准数据集(如ImageNet)上训练得到的模型权重来提高特定任务的效率。

为什么选择迁移学习

在深入了解迁移学习如何工作之前,先来看看进行迁移学习后能获得的好处。迁移学习过程中的学习是快速的——普通的卷积神经网络可能需要几天甚至几周来训练,但通过迁移学习可以缩短这个过程。迁移学习模型通常比定制模型的准确率高出20%,并且需要的训练数据更少——由于已经在大型数据集上训练过,模型已经能够检测到特定的特征,因此需要更少的训练数据来进一步改进模型。

图像数据的迁移学习

VGG架构

from torchvision import models, transforms import torch import torch.nn as nn import torch.optim as optim from torch.optim.lr_scheduler import StepLR import time import copy import numpy as np import matplotlib.pyplot as plt # 预处理 transform = transforms.Compose([ transforms.Resize((224, 224)), transforms.RandomHorizontalFlip(), transforms.ColorJitter(brightness=0.2, contrast=0.2, saturation=0.1, hue=0.1), transforms.RandomAffine(degrees=40, translate=None, scale=(1, 2), shear=15, resample=False, fillcolor=0), transforms.ToTensor(), transforms.Normalize((0.485, 0.456, 0.406), (0.229, 0.224, 0.225)) ]) # 加载数据集 from torchvision.datasets import ImageFolder train_dataset = ImageFolder('path_to_train_dataset', transform=transform) trainloader = torch.utils.data.DataLoader(train_dataset, batch_size=8, shuffle=True) # 可视化数据集 def imshow(inp, title=None): ""“Imshow 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 inputs, classes = next(iter(trainloader)) out = torchvision.utils.make_grid(inputs) imshow(out, title=[class_names[x] for x in classes])

导入和训练模型

device = torch.device("cuda" if torch.cuda.is_available() else "cpu") model_ft = models.vgg16(pretrained=True) # 划分数据集 num_ftrs = model_ft.fc.in_features model_ft.fc = nn.Linear(num_ftrs, 2) criterion = nn.CrossEntropyLoss() optimizer_ft = optim.SGD(model_ft.parameters(), lr=0.001, momentum=0.9) exp_lr_scheduler = lr_scheduler.StepLR(optimizer_ft, step_size=7, gamma=0.1) def train_model(model, criterion, optimizer, scheduler, num_epochs=25): since = time.time() best_model_wts = copy.deepcopy(model.state_dict()) best_acc = 0.0 for epoch in range(num_epochs): print('Epoch {}/{}'.format(epoch, num_epochs - 1)) print('-' * 10) 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 for inputs, labels in trainloader: inputs = inputs.to(device) labels = labels.to(device) optimizer.zero_grad() with torch.set_grad_enabled(phase == 'train'): outputs = model(inputs) _, preds = torch.max(outputs, 1) loss = criterion(outputs, labels) if phase == 'train': loss.backward() optimizer.step() 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 epoch_acc = running_corrects.double() / dataset_sizes print('{} Loss: {:.4f} Acc: {:.4f}'.format( phase, epoch_loss, epoch_acc)) print() time_elapsed = time.time() - since print('Training complete in {:.0f}m {:.0f}s'.format( time_elapsed // 60, time_elapsed % 60)) model.load_state_dict(best_model_wts) return model model_ft = train_model(model_ft, criterion, optimizer_ft, exp_lr_scheduler, num_epochs=25)
沪ICP备2024098111号-1
上海秋旦网络科技中心:上海市奉贤区金大公路8218号1幢 联系电话:17898875485