-
Notifications
You must be signed in to change notification settings - Fork 3
/
Copy pathdataset.py
148 lines (112 loc) · 4.45 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
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
140
141
142
143
144
145
146
147
148
import pickle
import socket
from PIL import Image
import glob
import torchvision
import numpy as np
import ipdb
import torch.nn.functional as F
st = ipdb.set_trace
import torch
import time
class ClevrDataset(torch.utils.data.Dataset):
def __init__(self, opt):
self.opt = opt
root_file = opt.root_folder
self.all_files = glob.glob(f'{root_file}/clevr_train/*')
self.resize = torchvision.transforms.Resize((opt.image_height,opt.image_width))
self.resize_mask = torchvision.transforms.Resize((opt.image_height,opt.image_width),torchvision.transforms.InterpolationMode.NEAREST)
def __len__(self):
return len(self.all_files)
def __getitem__(self, idx):
if self.opt.overfit:
idx = 0
file_val = self.all_files[idx]
pickled_file = pickle.load(open(file_val,'rb'))
rgb_val = torch.from_numpy(pickled_file['image']).squeeze().float()
images = rgb_val / 256.0
# Normalize to [0, 1] range.
gt_mask_val = torch.from_numpy(np.argmax(pickled_file['mask'].squeeze(),0))
images = images.permute(2,0,1).unsqueeze(0)
images = self.resize(images)
gt_mask_val = self.resize_mask(gt_mask_val.unsqueeze(0))
images = images.squeeze()
gt_mask_val = gt_mask_val.squeeze()
max_objs = gt_mask_val.max()
gt_indices = torch.zeros(self.opt.num_slots)
gt_indices[:max_objs] = 1.0
gt_mask_val = F.one_hot(gt_mask_val, self.opt.num_slots).permute(2,0,1)
return images, gt_mask_val , gt_indices
class ClevrTexDataset(torch.utils.data.Dataset):
def __init__(self, opt):
self.opt = opt
root_file = opt.root_folder
self.all_files = glob.glob(f'{root_file}/clevr_tex/*')
self.resize = torchvision.transforms.Resize((opt.image_height,opt.image_width))
self.resize_mask = torchvision.transforms.Resize((opt.image_height,opt.image_width),torchvision.transforms.InterpolationMode.NEAREST)
def __len__(self):
return len(self.all_files)
def __getitem__(self, idx):
if self.opt.overfit:
idx = 0
if self.opt.specific_example != 'None':
idx = int(self.opt.specific_example)
file_val = self.all_files[idx]
pickled_file = pickle.load(open(file_val,'rb'))
rgb_val = torch.from_numpy(pickled_file['image']).squeeze().float()
images = rgb_val / 256.0
# Normalize to [0, 1] range.
gt_mask_val = torch.from_numpy(np.argmax(pickled_file['mask'].squeeze(),0))
images = images.permute(2,0,1).unsqueeze(0)
images = self.resize(images)
gt_mask_val = self.resize_mask(gt_mask_val.unsqueeze(0))
images = images.squeeze()
gt_mask_val = gt_mask_val.squeeze()
max_objs = gt_mask_val.max()
gt_indices = torch.zeros(self.opt.num_slots)
gt_indices[:max_objs] = 1.0
gt_mask_val = F.one_hot(gt_mask_val, self.opt.num_slots).permute(2,0,1)
return images, gt_mask_val , gt_indices
def get_dataloader(opt, dataset):
# Improve reproducibility in dataloader.
data_loader = torch.utils.data.DataLoader(
dataset,
batch_size=opt.batch_size,
drop_last=True,
shuffle=True,
num_workers=opt.num_workers,
persistent_workers=opt.persistent_worker,
)
iterator = iter(data_loader)
return data_loader, iterator
def get_data(opt):
if opt.dataset_name == "clevr_tex":
dataset = ClevrTexDataset(opt)
else:
dataset = ClevrDataset(opt)
loader, iterator = get_dataloader(opt, dataset)
return loader, iterator
def get_data_tta(opt):
if opt.dataset_name == "clevr_tex":
dataset = ClevrTexDataset(opt)
else:
dataset = ClevrDataset(opt)
return dataset
def get_input(opt, iterator, train_loader):
time_init = time.time()
try:
input = next(iterator)
except StopIteration:
iterator = iter(train_loader)
input = next(iterator)
print("stop loading")
time_init = time.time()
image, gt_mask, gt_indices = input
image = image.to(opt.device)
gt_mask = gt_mask.to(opt.device)
gt_indices = gt_indices.to(opt.device)
feed_dict = {}
feed_dict["image"] = image
feed_dict["gt_mask"] = gt_mask
feed_dict["gt_indices"] = gt_indices
return feed_dict