def make_train_transform(self):
return transforms.Compose([
pt_util.ToPILImage(),
transforms.RandomResizedCrop(self.size, scale=(0.2, 1.0)),
transforms.RandomHorizontalFlip(),
pt_util.ToTensor(scale=255),
transforms.Normalize(mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
])
def make_val_transform(self):
return transforms.Compose([
pt_util.ToPILImage(),
transforms.Resize(
(int(self.size[0] / 0.875), int(self.size[1] / 0.875)),
interpolation=Image.BILINEAR),
transforms.CenterCrop(self.size),
pt_util.ToTensor(scale=255),
transforms.Normalize(mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
])
def make_train_transform(self):
return transforms.Compose([
transforms.RandomResizedCrop(self.size, scale=(0.08, 1.0)),
transforms.ColorJitter(0.4, 0.4, 0.4, 0.2),
# transforms.RandomResizedCrop(self.size, scale=(0.2, 1.0)),
transforms.RandomGrayscale(p=0.2),
# transforms.ColorJitter(0.4, 0.4, 0.4, 0.4),
transforms.RandomHorizontalFlip(),
pt_util.ToTensor(scale=255),
transforms.Normalize(mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
])
def __init__(self, args, data_subset="train"):
seqs = GOT10k(args.data_path, subset=data_subset, return_meta=True)
self.cfg = args.cfg
pair_transform = SiamFCTransforms(exemplar_sz=self.cfg["exemplar_sz"],
instance_sz=self.cfg["instance_sz"],
context=self.cfg["context"])
if data_subset == "train":
transform = transforms.Compose([
transforms.RandomApply([transforms.Lambda(fliplr)], 0.5),
pt_util.ToTensor(scale=255),
transforms.Normalize(mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
])
else:
transform = transforms.Compose([
pt_util.ToTensor(scale=255),
transforms.Normalize(mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
])
super(GOT10kDataset, self).__init__(args, seqs, data_subset,
pair_transform, transform)
def make_train_transform(self):
return transforms.Compose([
transforms.RandomResizedCrop(self.size, scale=(0.2, 1.0)),
transforms.RandomGrayscale(p=0.2),
transforms.ColorJitter(0.4, 0.4, 0.4, 0.4),
transforms.RandomHorizontalFlip(),
pt_util.ToTensor(scale=255),
transforms.Normalize(mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
transforms.RandomApply(
[util_functions.RandomGaussianBlur(self.size[0] // 10)],
p=0.5),
])
def make_train_transform(self):
return transforms.Compose([
pt_util.ToPILImage(),
transforms.RandomResizedCrop(self.size,
scale=(0.2, 1),
ratio=(0.7, 1.4),
interpolation=Image.BILINEAR),
transforms.ColorJitter(0.4, 0.4, 0.4, 0.2),
# transforms.RandomApply([transforms.ColorJitter(0.4, 0.4, 0.4, 0.4)], p=0.8),
transforms.RandomGrayscale(p=0.2),
transforms.RandomHorizontalFlip(),
pt_util.ToTensor(scale=255),
transforms.Normalize(mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
])