File size: 2,881 Bytes
178d33b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
import torchvision.transforms as tvs_trans

from openood.utils.config import Config

from .transform import Convert, interpolation_modes, normalization_dict


class RandAugmentPreprocessor():
    def __init__(self, config: Config):
        self.pre_size = config.dataset.pre_size
        self.image_size = config.dataset.image_size
        self.interpolation = interpolation_modes[config.dataset.interpolation]
        normalization_type = config.dataset.normalization_type
        if normalization_type in normalization_dict.keys():
            self.mean = normalization_dict[normalization_type][0]
            self.std = normalization_dict[normalization_type][1]
        else:
            self.mean = [0.5, 0.5, 0.5]
            self.std = [0.5, 0.5, 0.5]

        self.n = config.preprocessor.n
        self.m = config.preprocessor.m

        if 'imagenet' in config.dataset.name:
            self.transform = tvs_trans.Compose([
                tvs_trans.RandomResizedCrop(self.image_size,
                                            interpolation=self.interpolation),
                tvs_trans.RandomHorizontalFlip(0.5),
                tvs_trans.RandAugment(num_ops=self.n,
                                      magnitude=self.m,
                                      interpolation=self.interpolation),
                tvs_trans.ToTensor(),
                tvs_trans.Normalize(mean=self.mean, std=self.std),
            ])
        elif 'aircraft' in config.dataset.name or 'cub' in config.dataset.name:
            self.transform = tvs_trans.Compose([
                tvs_trans.Resize(self.pre_size,
                                 interpolation=self.interpolation),
                tvs_trans.RandomCrop(self.image_size),
                tvs_trans.RandomHorizontalFlip(),
                tvs_trans.RandAugment(num_ops=self.n,
                                      magnitude=self.m,
                                      interpolation=self.interpolation),
                tvs_trans.ToTensor(),
                tvs_trans.Normalize(mean=self.mean, std=self.std),
            ])
        else:
            self.transform = tvs_trans.Compose([
                Convert('RGB'),
                tvs_trans.RandAugment(num_ops=self.n,
                                      magnitude=self.m,
                                      interpolation=self.interpolation),
                tvs_trans.Resize(self.pre_size,
                                 interpolation=self.interpolation),
                tvs_trans.CenterCrop(self.image_size),
                tvs_trans.RandomHorizontalFlip(),
                tvs_trans.RandomCrop(self.image_size, padding=4),
                tvs_trans.ToTensor(),
                tvs_trans.Normalize(mean=self.mean, std=self.std),
            ])

    def setup(self, **kwargs):
        pass

    def __call__(self, image):
        return self.transform(image)