/
dataset.py
38 lines (28 loc) · 1.16 KB
/
dataset.py
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
import torch
import torchvision.transforms as transforms
from PIL import Image
from glob import glob
class Image2ImageDataSet(torch.utils.data.Dataset):
def __init__(self, seg_dir, real_dir, width, height, sample_size=1000):
self.transformation = transforms.Compose([
transforms.Resize((width, height), interpolation=Image.BICUBIC),
transforms.ToTensor(),
transforms.Normalize(mean=0.5, std=0.5)
])
seg_files = glob(seg_dir + '*.png')
real_files = glob(real_dir + '*.jpg')
self.seg_list = []
self.real_list = []
for image in range(sample_size):
self.seg_list.append(Image.open(seg_files[image]))
self.real_list.append(Image.open(real_files[image]))
def __len__(self):
return len(self.seg_list)
def __getitem__(self, idx):
if torch.is_tensor(idx):
idx = idx.tolist()
seg_image = self.seg_list[idx]
real_image = self.real_list[idx]
seg_image = self.transformation(seg_image)
real_image = self.transformation(real_image)
return seg_image, real_image