TorchVision: PyTorch的图像处理库+工具集

目录

• 1. torchvision.datasets
• 2. torchvision.io
• 3. torchvision.models
• 4. torchvision.transforms
  • 4.1. 针对 PIL.Image 的图像处理
  • 4.2. 针对张量 torch.*Tensor 的操作
  • 4.3. 数据格式转换操作
  • 4.4. 其他操作

---

官网入门教程 & API

这个包中有四个大类：

• torchvision.datasets
• torchvision.models
• torchvision.transforms
• torchvision.utils

1. torchvision.datasets

• CelebA
• CIFAR
• Cityscapes
• COCO
• DatasetFolder
• EMNIST
• FakeData
• Fashion-MNIST
• Flickr
• HMDB51
• ImageFolder
• ImageNet
• Kinetics-400
• KMNIST
• LSUN
• MNIST
• Omniglot
• PhotoTour
• Places365
• QMNIST
• SBD
• SBU
• STL10
• SVHN
• UCF101
• USPS
• VOC

torch.utils.data.DataLoader(coco_cap, batch_size=args.batchSize, shuffle=True, num_workers=args.nThreads)

2. torchvision.io

• Video
• Fine-grained video API
• Image

3. torchvision.models

• Classification
  • AlexNet
  • VGG
  • ResNet
  • SqueezeNet
  • DenseNet
  • Inception v3
  • GoogLeNet
  • ShuffleNet v2
  • MobileNet v2
  • ResNeXt
  • Wide ResNet
  • MNASNet
• Semantic Segmentation
  • FCN ResNet50
  • ResNet101
  • DeepLabV3 ResNet50
  • ResNet101
• Object Detection, Instance Segmentation and Person Keypoint Detection
• Faster R-CNN ResNet-50 FPN
• Mask R-CNN ResNet-50 FPN
• Video classification

载入模型示例

import torchvision.models as models

resnet18 = models.resnet18()
alexnet = models.alexnet()
vgg16 = models.vgg16()
squeezenet = models.squeezenet1_0()
densenet = models.densenet161()
inception = models.inception_v3()
googlenet = models.googlenet()
shufflenet = models.shufflenet_v2_x1_0()
mobilenet = models.mobilenet_v2()
resnext50_32x4d = models.resnext50_32x4d()
wide_resnet50_2 = models.wide_resnet50_2()
mnasnet = models.mnasnet1_0()

你也可以选择使用 pretrained=True 参数来加载预训练参数。

以下罗列了常见的几种模型的 Top-1/5 error ：ImageNet 1-crop error rates (224x224)

Network	Top-1 error	Top-5 error
AlexNet	43.45	20.91
VGG-16	28.41	9.62
VGG-19	27.62	9.12
VGG-16 with batch normalization	26.63	8.50
VGG-19 with batch normalization	25.76	8.15
ResNet-18	30.24	10.92
ResNet-34	26.70	8.58
ResNet-50	23.85	7.13
ResNet-101	22.63	6.44
ResNet-152	21.69	5.94
Densenet-161	22.35	6.20
MobileNet V2	28.12	9.71
ResNeXt-50-32x4d	22.38	6.30
ResNeXt-101-32x8d	20.69	5.47
Wide ResNet-50-2	21.49	5.91
Wide ResNet-101-2	21.16	5.72

4. torchvision.transforms

• Scriptable transforms
• Compositions of transforms
• Transforms on PIL Image and torch.*Tensor
• Transforms on PIL Image only
• Transforms on torch.*Tensor only
• Conversion Transforms
• Generic Transforms
• Functional Transforms

4.1. 针对 PIL.Image 的图像处理

• Scale(size, interpolation=Image.BILINEAR)
• CenterCrop(size) - center-crops the image to the given size
• RandomCrop(size, padding=0)
• RandomHorizontalFlip()
• RandomSizedCrop(size, interpolation=Image.BILINEAR)
• Pad(padding, fill=0)

4.2. 针对张量 torch.*Tensor 的操作

• Normalize(mean, std)

4.3. 数据格式转换操作

• ToTensor()

4.4. 其他操作

例如：

transforms.Lambda(lambda x: x.add(10))  # 将每个像素值加10