trainset = torchvision.datasets.CIFAR10(root=self._dataroot,
train=True,
download=True,
transform=transform_train)
if load_test:
testset = torchvision.datasets.CIFAR10(root=self._dataroot,
train=False,
download=True,
transform=transform_test)
return trainset, testset
@overrides
def get_transforms(self) -> tuple:
MEAN = [0.49139968, 0.48215827, 0.44653124]
STD = [0.24703233, 0.24348505, 0.26158768]
transf = [
transforms.RandomCrop(32, padding=4),
transforms.RandomHorizontalFlip()
]
normalize = [transforms.ToTensor(), transforms.Normalize(MEAN, STD)]
train_transform = transforms.Compose(transf + normalize)
test_transform = transforms.Compose(normalize)
return train_transform, test_transform
register_dataset_provider('cifar10', Cifar10Provider)
testpath = os.path.join(self._dataroot, 'food-101', 'test')
testset = torchvision.datasets.ImageFolder(
testpath, transform=transform_train)
return trainset, testset
@overrides
def get_transforms(self) -> tuple:
# TODO: Need to rethink the food101 transforms
MEAN = [0.5451, 0.4435, 0.3436]
STD = [0.2171, 0.2251,
0.2260] # TODO: should be [0.2517, 0.2521, 0.2573]
train_transf = [
transforms.Resize((32, 32)),
transforms.RandomCrop(32, padding=4),
transforms.RandomHorizontalFlip()
]
# food101 has images of varying sizes and are ~512 each side
test_transf = [transforms.Resize((32, 32))]
normalize = [transforms.ToTensor(), transforms.Normalize(MEAN, STD)]
train_transform = transforms.Compose(train_transf + normalize)
test_transform = transforms.Compose(test_transf + normalize)
return train_transform, test_transform
register_dataset_provider('food101', Food101Provider)
if load_train:
trainset = torchvision.datasets.FashionMNIST(root=self._dataroot,
train=True, download=True, transform=transform_train)
if load_test:
testset = torchvision.datasets.FashionMNIST(root=self._dataroot,
train=False, download=True, transform=transform_test)
return trainset, testset
@overrides
def get_transforms(self)->tuple:
MEAN = [0.28604063146254594]
STD = [0.35302426207299326]
transf = [
transforms.RandomAffine(degrees=15, translate=(0.1, 0.1),
scale=(0.9, 1.1), shear=0.1),
transforms.RandomVerticalFlip()
]
normalize = [
transforms.ToTensor(),
transforms.Normalize(MEAN, STD)
]
train_transform = transforms.Compose(transf + normalize)
test_transform = transforms.Compose(normalize)
return train_transform, test_transform
register_dataset_provider('fashion_mnist', FashionMnistProvider)
download=True,
transform=transform_train)
if load_test:
testset = torchvision.datasets.MNIST(root=self._dataroot,
train=False,
download=True,
transform=transform_test)
return trainset, testset
@overrides
def get_transforms(self) -> tuple:
MEAN = [0.13066051707548254]
STD = [0.30810780244715075]
transf = [
transforms.RandomAffine(degrees=15,
translate=(0.1, 0.1),
scale=(0.9, 1.1),
shear=0.1)
]
normalize = [transforms.ToTensor(), transforms.Normalize(MEAN, STD)]
train_transform = transforms.Compose(transf + normalize)
test_transform = transforms.Compose(normalize)
return train_transform, test_transform
register_dataset_provider('mnist', MnistProvider)
return trainset, testset
@overrides
def get_transforms(self) -> tuple:
# MEAN, STD computed for mit67
MEAN = [0.4893, 0.4270, 0.3625]
STD = [0.2631, 0.2565, 0.2582]
# transformations match that in
# https://github.com/antoyang/NAS-Benchmark/blob/master/DARTS/preproc.py
train_transf = [
transforms.RandomResizedCrop(224),
transforms.RandomHorizontalFlip(),
transforms.ColorJitter(brightness=0.4,
contrast=0.4,
saturation=0.4,
hue=0.2)
]
test_transf = [transforms.Resize(256), transforms.CenterCrop(224)]
normalize = [transforms.ToTensor(), transforms.Normalize(MEAN, STD)]
train_transform = transforms.Compose(train_transf + normalize)
test_transform = transforms.Compose(test_transf + normalize)
return train_transform, test_transform
register_dataset_provider('mit67', Mit67Provider)
def get_transforms(self)->tuple:
# MEAN, STD computed for sport8
MEAN = [0.4734, 0.4856, 0.4526]
STD = [0.2478, 0.2444, 0.2667]
# transformations match that in
# https://github.com/antoyang/NAS-Benchmark/blob/master/DARTS/preproc.py
train_transf = [
transforms.RandomResizedCrop(224),
transforms.RandomHorizontalFlip(),
transforms.ColorJitter(
brightness=0.4,
contrast=0.4,
saturation=0.4,
hue=0.2)
]
test_transf = [transforms.Resize(256), transforms.CenterCrop(224)]
normalize = [
transforms.ToTensor(),
transforms.Normalize(MEAN, STD)
]
train_transform = transforms.Compose(train_transf + normalize)
test_transform = transforms.Compose(test_transf + normalize)
return train_transform, test_transform
register_dataset_provider('sport8', Sport8Provider)
split='extra',
download=True,
transform=transform_train)
trainset = ConcatDataset([trainset, extraset])
if load_test:
testset = torchvision.datasets.SVHN(root=self._dataroot,
split='test',
download=True,
transform=transform_test)
return trainset, testset
@overrides
def get_transforms(self) -> tuple:
MEAN = [0.4914, 0.4822, 0.4465]
STD = [0.2023, 0.1994, 0.20100]
transf = [
transforms.RandomCrop(32, padding=4),
transforms.RandomHorizontalFlip()
]
normalize = [transforms.ToTensor(), transforms.Normalize(MEAN, STD)]
train_transform = transforms.Compose(transf + normalize)
test_transform = transforms.Compose(normalize)
return train_transform, test_transform
register_dataset_provider('svhn', SvhnProvider)
return trainset, testset
@overrides
def get_transforms(self) -> tuple:
# MEAN, STD computed for flower102
MEAN = [0.5190, 0.4101, 0.3274]
STD = [0.2972, 0.2488, 0.2847]
# transformations match that in
# https://github.com/antoyang/NAS-Benchmark/blob/master/DARTS/preproc.py
train_transf = [
transforms.RandomResizedCrop(224),
transforms.RandomHorizontalFlip(),
transforms.ColorJitter(brightness=0.4,
contrast=0.4,
saturation=0.4,
hue=0.2)
]
test_transf = [transforms.Resize(256), transforms.CenterCrop(224)]
normalize = [transforms.ToTensor(), transforms.Normalize(MEAN, STD)]
train_transform = transforms.Compose(train_transf + normalize)
test_transform = transforms.Compose(test_transf + normalize)
return train_transform, test_transform
register_dataset_provider('flower102', Flower102Provider)
transform_train = transforms.Compose([
transforms.RandomResizedCrop(
224,
scale=(
0.08,
1.0), # TODO: these two params are normally not specified
interpolation=Image.BICUBIC),
transforms.RandomHorizontalFlip(),
transforms.ColorJitter(brightness=0.4,
contrast=0.4,
saturation=0.4,
hue=0.2),
transforms.ToTensor(),
# TODO: Lighting is not used in original darts paper
# Lighting(0.1, _IMAGENET_PCA['eigval'], _IMAGENET_PCA['eigvec']),
transforms.Normalize(mean=MEAN, std=STD)
])
transform_test = transforms.Compose([
transforms.Resize(256),
transforms.CenterCrop(224),
transforms.ToTensor(),
transforms.Normalize(mean=MEAN, std=STD)
])
return transform_train, transform_test
register_dataset_provider('imagenet', ImagenetProvider)
