FakeImageDetection/dataset.py at master · Njoom/FakeImageDetection · GitHub

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
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
import torch
from torchvision.io import read_image
from torchvision import transforms
from torch.utils.data import Dataset
from typing import Any, Callable, Optional
import os
from PIL import Image
import numpy as np
from tqdm import tqdm
import random

class ForenSynths(Dataset):
    def __init__(self, root_dir, transform=None):
        self.root_dir = root_dir
        self.transform = transform
        self.classes = ['0_real', '1_fake']
        self.data = []

        # Iterate over the categories
        for category in os.listdir(root_dir):
            category_path = os.path.join(root_dir, category)

            # Iterate over class names (real/fake)
            for class_name in self.classes:
                class_path = os.path.join(category_path, class_name)

                # Iterate over files
                for file_name in os.listdir(class_path):
                    file_path = os.path.join(class_path, file_name)

                    # Append a tuple (file_path, class_index)
                    self.data.append((file_path, self.classes.index(class_name)))

    def __len__(self):
        return len(self.data)

    def __getitem__(self, index):
        img_path, label = self.data[index]

        # Load image
        image = Image.open(img_path).convert("RGB")

        # Apply transforms if any
        if self.transform:
            image = self.transform(image)

        return image, label

class OjhaCVPR23(Dataset):
    def __init__(self, root_dir, transform=None):
        self.transform = transform
        self.data = []
        self.fake_index = 1  # Index of 'fake' in ['real', 'fake']
        self.real_index = 0  # Index of 'real' in ['real', 'fake']

        sub_folders = os.listdir(root_dir)

        if '1_fake' in sub_folders and '0_real' in sub_folders:
            # This is the 'biggan' case
            fake_dir = os.path.join(root_dir, '1_fake')
            real_dir = os.path.join(root_dir, '0_real')
            self._process_folder(fake_dir, self.fake_index)
            self._process_folder(real_dir, self.real_index)
        else:
            # This is the 'cyclegan' case
            for folder in sub_folders:
                fake_dir = os.path.join(root_dir, folder, '1_fake')
                real_dir = os.path.join(root_dir, folder, '0_real')
                self._process_folder(fake_dir, self.fake_index)
                self._process_folder(real_dir, self.real_index)

    def _process_folder(self, folder_path, index):
        # Iterate over files
        for file_name in os.listdir(folder_path):
            file_path = os.path.join(folder_path, file_name)

            # Append a tuple (file_path, index)
            self.data.append((file_path, index))

    def __len__(self):
        return len(self.data)

    def __getitem__(self, index):
        img_path, label = self.data[index]

        # Load image
        image = Image.open(img_path).convert("RGB")

        # Apply transforms if any
        if self.transform:
            image = self.transform(image)

        return image, label

class Wang_CVPR20(Dataset):
    def __init__(self, root_dir, transform=None):
        self.transform = transform
        self.data = []
        self.fake_index = 1  # Index of 'fake' in ['real', 'fake']
        self.real_index = 0  # Index of 'real' in ['real', 'fake']

        sub_folders = os.listdir(root_dir)

        if '1_fake' in sub_folders and '0_real' in sub_folders:
            # This is the 'biggan' case
            fake_dir = os.path.join(root_dir, '1_fake')
            real_dir = os.path.join(root_dir, '0_real')
            self._process_folder(fake_dir, self.fake_index)
            self._process_folder(real_dir, self.real_index)
        else:
            # This is the 'cyclegan' case
            for folder in sub_folders:
                fake_dir = os.path.join(root_dir, folder, '1_fake')
                real_dir = os.path.join(root_dir, folder, '0_real')
                self._process_folder(fake_dir, self.fake_index)
                self._process_folder(real_dir, self.real_index)

    def _process_folder(self, folder_path, index):
        # Iterate over files
        for file_name in os.listdir(folder_path):
            file_path = os.path.join(folder_path, file_name)

            # Append a tuple (file_path, index)
            self.data.append((file_path, index))

    def __len__(self):
        return len(self.data)

    def __getitem__(self, index):
        img_path, label = self.data[index]

        # Load image
        image = Image.open(img_path).convert("RGB")

        # Apply transforms if any
        if self.transform:
            image = self.transform(image)

        return image, label