def get_dataset(opts):
""" Dataset And Augmentation
"""
if opts.crop_val:
val_transform = et.ExtCompose([
et.ExtResize(opts.crop_size),
et.ExtCenterCrop(opts.crop_size),
et.ExtToTensor(),
et.ExtNormalize(mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
])
else:
val_transform = et.ExtCompose([
et.ExtToTensor(),
et.ExtNormalize(mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
])
val_dst = VOCSegmentation(root=opts.data_root,
year=opts.year,
image_set='val',
download=False,
transform=val_transform,
ret_fname=True)
return val_dst
def get_dataset(opts):
""" Dataset And Augmentation
"""
if opts.dataset == 'voc':
train_transform = et.ExtCompose([
#et.ExtResize(size=opts.crop_size),
et.ExtRandomScale((0.5, 2.0)),
et.ExtRandomCrop(size=(opts.crop_size, opts.crop_size), pad_if_needed=True),
et.ExtRandomHorizontalFlip(),
et.ExtToTensor(),
et.ExtNormalize(mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
])
if opts.crop_val:
val_transform = et.ExtCompose([
et.ExtResize(opts.crop_size),
et.ExtCenterCrop(opts.crop_size),
et.ExtToTensor(),
et.ExtNormalize(mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
])
else:
val_transform = et.ExtCompose([
et.ExtToTensor(),
et.ExtNormalize(mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
])
train_dst = VOCSegmentation(root=opts.data_root, year=opts.year,
image_set='train', download=opts.download, transform=train_transform)
val_dst = VOCSegmentation(root=opts.data_root, year=opts.year,
image_set='val', download=False, transform=val_transform)
if opts.dataset == 'cityscapes':
train_transform = et.ExtCompose([
#et.ExtResize( 512 ),
et.ExtRandomCrop(size=(opts.crop_size, opts.crop_size)),
et.ExtColorJitter( brightness=0.5, contrast=0.5, saturation=0.5 ),
et.ExtRandomHorizontalFlip(),
et.ExtToTensor(),
et.ExtNormalize(mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
])
val_transform = et.ExtCompose([
et.ExtResize( 256 ),
et.ExtToTensor(),
et.ExtNormalize(mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
])
train_dst = Cityscapes(root=opts.data_root,
split='train', transform=train_transform)
val_dst = Cityscapes(root=opts.data_root,
split='val', transform=val_transform)
return train_dst, val_dst
def get_test_dataset(opts):
""" Dataset And Augmentation
"""
if opts.crop_val:
val_transform = et.ExtCompose([
et.ExtResize(opts.crop_size),
et.ExtCenterCrop(opts.crop_size),
et.ExtToTensor(),
et.ExtNormalize(mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
])
else:
val_transform = et.ExtCompose([
et.ExtToTensor(),
et.ExtNormalize(mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
])
# 生成test_dir
test_dst = DataSegmentationTest(transform=val_transform,
test_dir=opts.test_dir,
png_dir=opts.png_dir)
return test_dst
def get_dataset(opts):
""" Dataset And Augmentation
"""
train_transform = et.ExtCompose([
#et.ExtResize(size=opts.crop_size),
et.ExtRandomScale((0.5, 2.0)),
et.ExtRandomCrop(size=(opts.crop_size, opts.crop_size),
pad_if_needed=True),
et.ExtRandomHorizontalFlip(),
et.ExtToTensor(),
et.ExtNormalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]),
])
if opts.crop_val:
val_transform = et.ExtCompose([
et.ExtResize(opts.crop_size),
et.ExtCenterCrop(opts.crop_size),
et.ExtToTensor(),
et.ExtNormalize(mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
])
else:
val_transform = et.ExtCompose([
et.ExtToTensor(),
et.ExtNormalize(mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
])
train_dst = DataSegmentation(image_set='train',
transform=train_transform,
jpg_dir=opts.jpg_dir,
png_dir=opts.png_dir,
list_dir=opts.list_dir)
val_dst = DataSegmentation(image_set='val',
transform=val_transform,
jpg_dir=opts.jpg_dir,
png_dir=opts.png_dir,
list_dir=opts.list_dir)
return train_dst, val_dst
def get_dataloader(args):
if args.dataset.lower()=='mnist':
train_loader = torch.utils.data.DataLoader(
datasets.MNIST(args.data_root, train=True, download=True,
transform=transforms.Compose([
transforms.Resize((32, 32)),
transforms.ToTensor(),
transforms.Normalize((0.1307,), (0.3081,))
])),
batch_size=args.batch_size, shuffle=True, num_workers=2)
test_loader = torch.utils.data.DataLoader(
datasets.MNIST(args.data_root, train=False, download=True,
transform=transforms.Compose([
transforms.Resize((32, 32)),
transforms.ToTensor(),
transforms.Normalize((0.1307,), (0.3081,))
])),
batch_size=args.batch_size, shuffle=True, num_workers=2)
elif args.dataset.lower()=='cifar10':
train_loader = torch.utils.data.DataLoader(
datasets.CIFAR10(args.data_root, train=True, download=True,
transform=transforms.Compose([
transforms.RandomCrop(32, padding=4),
transforms.RandomHorizontalFlip(),
transforms.ToTensor(),
transforms.Normalize((0.4914, 0.4822, 0.4465), (0.2023, 0.1994, 0.2010)),
])),
batch_size=args.batch_size, shuffle=True, num_workers=2)
test_loader = torch.utils.data.DataLoader(
datasets.CIFAR10(args.data_root, train=False, download=True,
transform=transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.4914, 0.4822, 0.4465), (0.2023, 0.1994, 0.2010)),
])),
batch_size=args.batch_size, shuffle=True, num_workers=2)
elif args.dataset.lower()=='cifar100':
train_loader = torch.utils.data.DataLoader(
datasets.CIFAR100(args.data_root, train=True, download=True,
transform=transforms.Compose([
transforms.RandomCrop(32, padding=4),
transforms.RandomHorizontalFlip(),
transforms.ToTensor(),
transforms.Normalize((0.4914, 0.4822, 0.4465), (0.2023, 0.1994, 0.2010)),
])),
batch_size=args.batch_size, shuffle=True, num_workers=2)
test_loader = torch.utils.data.DataLoader(
datasets.CIFAR100(args.data_root, train=False, download=True,
transform=transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.4914, 0.4822, 0.4465), (0.2023, 0.1994, 0.2010)),
])),
batch_size=args.batch_size, shuffle=True, num_workers=2)
elif args.dataset.lower()=='caltech101':
train_loader = torch.utils.data.DataLoader(
Caltech101(args.data_root, train=True, download=args.download,
transform=transforms.Compose([
transforms.Resize(128),
transforms.RandomCrop(128),
transforms.RandomHorizontalFlip(),
transforms.ToTensor(),
transforms.Normalize((0.5,), (0.5,))
])),
batch_size=args.batch_size, shuffle=True, num_workers=2)
test_loader = torch.utils.data.DataLoader(
Caltech101(args.data_root, train=False, download=args.download,
transform=transforms.Compose([
transforms.Resize(128),
transforms.CenterCrop(128),
transforms.ToTensor(),
transforms.Normalize((0.5,), (0.5,))
])),
batch_size=args.test_batch_size, shuffle=False, num_workers=2)
############ Segmentation
elif args.dataset.lower()=='camvid':
train_loader = torch.utils.data.DataLoader(
CamVid(args.data_root, split='train',
transform=ext_transforms.ExtCompose([
ext_transforms.ExtResize(256),
ext_transforms.ExtRandomCrop(128, pad_if_needed=True),
ext_transforms.ExtRandomHorizontalFlip(),
ext_transforms.ExtToTensor(),
ext_transforms.ExtNormalize((0.5,), (0.5,))
])),
batch_size=args.batch_size, shuffle=True, num_workers=2)
test_loader = torch.utils.data.DataLoader(
CamVid(args.data_root, split='test',
transform=ext_transforms.ExtCompose([
ext_transforms.ExtResize(256),
ext_transforms.ExtToTensor(),
ext_transforms.ExtNormalize((0.5,), (0.5,))
])),
batch_size=args.test_batch_size, shuffle=False, num_workers=2)
elif args.dataset.lower() in ['nyuv2']:
train_loader = torch.utils.data.DataLoader(
NYUv2(args.data_root, split='train',
transform=ext_transforms.ExtCompose([
ext_transforms.ExtResize(256),
ext_transforms.ExtRandomCrop(128, pad_if_needed=True),
ext_transforms.ExtRandomHorizontalFlip(),
ext_transforms.ExtToTensor(),
ext_transforms.ExtNormalize((0.5,), (0.5,))
])),
batch_size=args.batch_size, shuffle=True, num_workers=2)
test_loader = torch.utils.data.DataLoader(
NYUv2(args.data_root, split='test',
transform=ext_transforms.ExtCompose([
ext_transforms.ExtResize(256),
ext_transforms.ExtToTensor(),
ext_transforms.ExtNormalize((0.5,), (0.5,))
])),
batch_size=args.test_batch_size, shuffle=False, num_workers=2)
return train_loader, test_loader
def get_dataset(opts):
""" Dataset And Augmentation
"""
if opts.dataset == 'voc':
train_transform = et.ExtCompose([
#et.ExtResize(size=opts.crop_size),
et.ExtRandomScale((0.5, 2.0)),
et.ExtRandomCrop(size=(opts.crop_size, opts.crop_size),
pad_if_needed=True),
et.ExtRandomHorizontalFlip(),
et.ExtToTensor(),
et.ExtNormalize(mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
])
if opts.crop_val:
val_transform = et.ExtCompose([
et.ExtResize(opts.crop_size),
et.ExtCenterCrop(opts.crop_size),
et.ExtToTensor(),
et.ExtNormalize(mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
])
else:
val_transform = et.ExtCompose([
et.ExtToTensor(),
et.ExtNormalize(mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
])
train_dst = VOCSegmentation(root=opts.data_root,
year=opts.year,
image_set='train',
download=opts.download,
transform=train_transform)
val_dst = VOCSegmentation(root=opts.data_root,
year=opts.year,
image_set='val',
download=False,
transform=val_transform)
if opts.dataset == 'cityscapes':
train_transform = et.ExtCompose([
#et.ExtResize( 512 ),
et.ExtRandomCrop(size=(opts.crop_size, opts.crop_size)),
et.ExtColorJitter(brightness=0.5, contrast=0.5, saturation=0.5),
et.ExtRandomHorizontalFlip(),
et.ExtToTensor(),
et.ExtNormalize(mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
])
val_transform = et.ExtCompose([
#et.ExtResize( 512 ),
et.ExtToTensor(),
et.ExtNormalize(mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
])
train_dst = Cityscapes(root=opts.data_root,
split='train',
transform=train_transform)
val_dst = Cityscapes(root=opts.data_root,
split='val',
transform=val_transform)
if opts.dataset == 'weedcluster':
train_dst = WeedClusterDataset(root=opts.data_root, split='train')
val_dst = WeedClusterDataset(root=opts.data_root, split='val')
if opts.dataset == 'cloudshadow':
train_dst = CloudShadowDataset(root=opts.data_root, split='train')
val_dst = CloudShadowDataset(root=opts.data_root, split='val')
if opts.dataset == 'doubleplant':
train_dst = DoublePlantDataset(root=opts.data_root, split='train')
val_dst = DoublePlantDataset(root=opts.data_root, split='val')
if opts.dataset == 'planterskip':
train_dst = PlanterSkipDataset(root=opts.data_root, split='train')
val_dst = PlanterSkipDataset(root=opts.data_root, split='val')
if opts.dataset == 'standingwater':
train_dst = StandingWaterDataset(root=opts.data_root, split='train')
val_dst = StandingWaterDataset(root=opts.data_root, split='val')
if opts.dataset == 'waterway':
train_dst = WaterwayDataset(root=opts.data_root, split='train')
val_dst = WaterwayDataset(root=opts.data_root, split='val')
return train_dst, val_dst
import torch
import torch.nn as nn
import torch.nn.functional as F
from PIL import Image
import matplotlib
import matplotlib.pyplot as plt
data_root = '/data/kdn/Dataset/Cityscapes'
crop_size = 520
train_transform = et.ExtCompose([
et.ExtRandomCrop(size=(crop_size, crop_size)),
et.ExtColorJitter(brightness=0.5, contrast=0.5, saturation=0.5),
et.ExtRandomHorizontalFlip(),
et.ExtToTensor(),
et.ExtNormalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]),
])
val_transform = et.ExtCompose([
et.ExtToTensor(),
et.ExtNormalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]),
])
train_dst = Cityscapes(root=data_root,
split='train',
transform=train_transform)
val_dst = Cityscapes(root=data_root, split='val', transform=val_transform)
train_loader = data.DataLoader(train_dst,
batch_size=2,
