def get_setting(args):
kwargs = {'num_workers': 1, 'pin_memory': True} if args.cuda else {}
path = os.path.join(args.data_folder)
if args.dataset == 'mnist':
num_class = 10
train_loader = torch.utils.data.DataLoader(datasets.MNIST(
path,
train=True,
download=True,
transform=transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.1307, ), (0.3081, ))
])),
batch_size=args.batch_size,
shuffle=True,
**kwargs)
test_loader = torch.utils.data.DataLoader(
datasets.MNIST(path,
train=False,
transform=transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.1307, ), (0.3081, ))
])),
batch_size=args.test_batch_size,
shuffle=True,
**kwargs)
elif args.dataset == 'smallNORB':
num_class = 5
train_loader = torch.utils.data.DataLoader(smallNORB(
path,
train=True,
download=True,
transform=transforms.Compose([
transforms.Resize(48),
transforms.RandomCrop(32),
transforms.ColorJitter(brightness=32. / 255, contrast=0.5),
transforms.ToTensor()
])),
batch_size=args.batch_size,
shuffle=True,
**kwargs)
test_loader = torch.utils.data.DataLoader(
smallNORB(path,
train=False,
transform=transforms.Compose([
transforms.Resize(48),
transforms.CenterCrop(32),
transforms.ToTensor()
])),
batch_size=args.test_batch_size,
shuffle=True,
**kwargs)
else:
raise NameError('Undefined dataset {}'.format(args.dataset))
return num_class, train_loader, test_loader
def get_setting(args):
kwargs = {'num_workers': 1, 'pin_memory': True} if args.cuda else {}
path = os.path.join(args.data_folder, args.dataset)
if args.dataset == 'mnist':
num_class = 10
train_loader = torch.utils.data.DataLoader(datasets.MNIST(
path,
train=True,
download=True,
transform=transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.1307, ), (0.3081, ))
])),
batch_size=args.batch_size,
shuffle=True,
**kwargs)
test_loader = torch.utils.data.DataLoader(
datasets.MNIST(path,
train=False,
transform=transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.1307, ), (0.3081, ))
])),
batch_size=args.test_batch_size,
shuffle=True,
**kwargs)
elif args.dataset == 'smallNORB':
num_class = 5
train_loader = torch.utils.data.DataLoader(smallNORB(
path,
train=True,
download=True,
transform=transforms.Compose([
transforms.Resize(48),
transforms.RandomCrop(32),
transforms.ColorJitter(brightness=32. / 255, contrast=0.5),
transforms.ToTensor()
])),
batch_size=args.batch_size,
shuffle=True,
**kwargs)
test_loader = torch.utils.data.DataLoader(
smallNORB(path,
train=False,
transform=transforms.Compose([
transforms.Resize(48),
transforms.CenterCrop(32),
transforms.ToTensor()
])),
batch_size=args.test_batch_size,
shuffle=True,
**kwargs)
elif args.dataset == 'gtrsb':
num_class = 43
full_dataset = GTRSB(path,
download=True,
transform=transforms.Compose([
transforms.Grayscale(),
transforms.Resize(
(48, 48), interpolation=Image.LANCZOS),
transforms.ToTensor()
]))
train_size = 39209
val_size = 12630
print(f"Train Size: {str(train_size)}")
print(f"Val Size: {str(val_size)}")
train_dataset, val_dataset = torch.utils.data.random_split(
full_dataset, [train_size, val_size])
train_loader = torch.utils.data.DataLoader(train_dataset,
batch_size=args.batch_size,
shuffle=True,
**kwargs)
val_loader = torch.utils.data.DataLoader(
val_dataset,
batch_size=args.test_batch_size,
shuffle=True,
**kwargs)
else:
raise NameError('Undefined dataset {}'.format(args.dataset))
return num_class, train_loader, val_loader
logger = util.statNothing()
elif args.dataset[:5] == 'MNIST':
train_dataset = datasets.MNIST(root='../../data/',
train=True,
transform=transforms.ToTensor(),
download=True)
test_dataset = datasets.MNIST(root='../../data/',
train=False,
transform=transforms.ToTensor())
logger = util.statClassification(args)
elif args.dataset == 'smallNORB': # transforms.Resize(48),
train_dataset = smallNORB('../../data/smallnorb/',
train=True,
download=True,
transform=transforms.Compose([
transforms.RandomCrop(64),
transforms.ColorJitter(brightness=32. /
255,
contrast=0.5),
transforms.ToTensor()
]))
test_dataset = smallNORB('../../data/smallnorb/',
train=False,
transform=transforms.Compose([
transforms.CenterCrop(64),
transforms.ToTensor()
]))
elif args.dataset == 'msra':
train_dataset = MARAHandDataset('../../data/cvpr15_MSRAHandGestureDB',
'train', 2)
#train_dataset.test()
test_dataset = MARAHandDataset('../../data/cvpr15_MSRAHandGestureDB',
def get_setting(args):
kwargs = {'num_workers': 1, 'pin_memory': True} if args.cuda else {}
path = os.path.join(args.data_folder, args.dataset)
if args.dataset == 'mnist':
num_class = 10
train_loader = torch.utils.data.DataLoader(
datasets.MNIST(path, train=True, download=True,
transform=transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.1307,), (0.3081,))
])),
batch_size=args.batch_size, shuffle=True, **kwargs)
test_loader = torch.utils.data.DataLoader(
datasets.MNIST(path, train=False,
transform=transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.1307,), (0.3081,))
])),
batch_size=args.test_batch_size, shuffle=True, **kwargs)
elif args.dataset == 'smallNORB':
num_class = 5
train_loader = torch.utils.data.DataLoader(
smallNORB(path, train=True, download=True,
transform=transforms.Compose([
transforms.Resize(48),
transforms.RandomCrop(32),
transforms.ColorJitter(brightness=32./255, contrast=0.5),
transforms.ToTensor()
])),
batch_size=args.batch_size, shuffle=True, **kwargs)
test_loader = torch.utils.data.DataLoader(
smallNORB(path, train=False,
transform=transforms.Compose([
transforms.Resize(48),
transforms.CenterCrop(32),
transforms.ToTensor()
])),
batch_size=args.test_batch_size, shuffle=True, **kwargs)
elif args.dataset == "KTH":
print("Preparing KTH data...")
print("train batch size: ", args.batch_size)
print("test batch size: ", args.test_batch_size)
train_data = prepareDataset("/Volumes/E_128/train")
test_data = prepareDataset("/Volumes/E_128/test")
num_class = 6
# train
classes={
"handwaving": 0,
"handclapping": 1,
"boxing": 2,
"walking": 3,
"running": 4,
"jogging": 5
}
action_dataset_train = ActionDataset(train_data, classes, transforms=transforms.Compose([
transforms.Resize(48),
transforms.RandomCrop(32),
transforms.ColorJitter(brightness=32./255, contrast=0.5),
transforms.ToTensor()
]))
train_loader = torch.utils.data.DataLoader(action_dataset_train, batch_size=args.batch_size,
shuffle=True, **kwargs)
action_dataset_test = ActionDataset(test_data, classes, transforms=transforms.Compose([
transforms.Resize(48),
transforms.RandomCrop(32),
transforms.ToTensor()
]))
test_loader = torch.utils.data.DataLoader(action_dataset_test, batch_size=args.test_batch_size,
shuffle=True, **kwargs)
else:
raise NameError('Undefined dataset {}'.format(args.dataset))
return num_class, train_loader, test_loader
def get_setting(args):
kwargs = {'num_workers': 1, 'pin_memory': True} if args.cuda else {}
path = os.path.join(args.data_folder, args.dataset)
# normalize = transforms.Normalize(mean=[x/255.0 for x in [125.3, 123.0, 113.9]],
# std=[x/255.0 for x in [63.0, 62.1, 66.7]])
transform_train = transforms.Compose([
transforms.Grayscale(num_output_channels=1),
transforms.RandomCrop(28, padding='valid'),
transforms.RandomHorizontalFlip(),
transforms.ToTensor()
])
transform_test = transforms.Compose(
[transforms.Grayscale(num_output_channels=1),
transforms.ToTensor()])
if args.dataset == 'mnist':
num_class = 10
train_loader = torch.utils.data.DataLoader(datasets.MNIST(
path,
train=True,
download=True,
transform=transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.1307, ), (0.3081, ))
])),
batch_size=args.batch_size,
shuffle=True,
**kwargs)
test_loader = torch.utils.data.DataLoader(
datasets.MNIST(path,
train=False,
transform=transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.1307, ), (0.3081, ))
])),
batch_size=args.test_batch_size,
shuffle=True,
**kwargs)
elif args.dataset == 'smallNORB':
num_class = 5
train_loader = torch.utils.data.DataLoader(smallNORB(
path,
train=True,
download=True,
transform=transforms.Compose([
transforms.Resize(48),
transforms.RandomCrop(32),
transforms.ColorJitter(brightness=32. / 255, contrast=0.5),
transforms.ToTensor()
])),
batch_size=args.batch_size,
shuffle=True,
**kwargs)
test_loader = torch.utils.data.DataLoader(
smallNORB(path,
train=False,
transform=transforms.Compose([
transforms.Resize(48),
transforms.CenterCrop(32),
transforms.ToTensor()
])),
batch_size=args.test_batch_size,
shuffle=True,
**kwargs)
elif args.dataset == 'cifar10':
num_class = 10
train_loader = torch.utils.data.DataLoader(datasets.CIFAR10(
path,
train=True,
download=True,
transform=transforms.Compose([
transforms.Grayscale(num_output_channels=1),
transforms.RandomCrop(32),
transforms.ToTensor()
])),
batch_size=args.batch_size,
shuffle=True,
**kwargs)
test_loader = torch.utils.data.DataLoader(
datasets.CIFAR10(path,
train=False,
transform=transforms.Compose([
transforms.Grayscale(num_output_channels=1),
transforms.ToTensor()
])),
batch_size=args.test_batch_size,
shuffle=True,
**kwargs)
else:
raise NameError('Undefined dataset {}'.format(args.dataset))
return num_class, train_loader, test_loader
