-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathdataset.py
288 lines (249 loc) · 13.7 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
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
from PIL import Image
from torchvision.transforms.functional import to_tensor
import torch.utils.data as data
import os
from glob import glob
from random import randint
import torch
import numpy as np
import cv2
from utils import video2frames
class Vimeo90k(data.Dataset):
def __init__(self, path, is_train=True, crop_size=None):
super().__init__()
train_list = os.path.join(path, 'tri_trainlist.txt')
test_list = os.path.join(path, 'tri_testlist.txt')
self.frame_list = os.path.join(path, 'sequences')
self.is_train = is_train
if self.is_train:
triplet_list = train_list
else:
triplet_list = test_list
with open(triplet_list) as triplet_list_file:
triplet_list = triplet_list_file.readlines()
triplet_list_file.close()
self.triplet_list = triplet_list[:-1]
if crop_size is None or crop_size <= 0:
self.crop_size = None
else:
if type(crop_size) is not tuple:
crop_size = (crop_size, crop_size)
self.crop_size = crop_size
def __len__(self):
return len(self.triplet_list)
def __getitem__(self, idx):
triplet_path = self.triplet_list[idx]
if triplet_path[-1:] == '\n':
triplet_path = triplet_path[:-1]
try:
vid_no, seq_no = triplet_path.split('/')
name = vid_no + '_' + seq_no
except:
name = triplet_path
triplet_path = os.path.join(self.frame_list, triplet_path)
im1 = os.path.join(triplet_path, 'im1.png')
im2 = os.path.join(triplet_path, 'im2.png')
im3 = os.path.join(triplet_path, 'im3.png')
# read image file
im1 = Image.open(im1).convert('RGB')
im2 = Image.open(im2).convert('RGB')
im3 = Image.open(im3).convert('RGB')
# data augmentation - random flip / sequence flip
if self.is_train:
h_flip_flag = randint(0, 1)
v_flip_flag = randint(0, 1)
if h_flip_flag == 1:
im1 = im1.transpose(Image.FLIP_LEFT_RIGHT)
im2 = im2.transpose(Image.FLIP_LEFT_RIGHT)
im3 = im3.transpose(Image.FLIP_LEFT_RIGHT)
if v_flip_flag == 1:
im1 = im1.transpose(Image.FLIP_TOP_BOTTOM)
im2 = im2.transpose(Image.FLIP_TOP_BOTTOM)
im3 = im3.transpose(Image.FLIP_TOP_BOTTOM)
order_reverse = randint(0, 1)
if order_reverse == 1:
tmp = im1
im1 = im3
im3 = tmp
if self.crop_size is not None:
# random crop
frame_w, frame_h = im1.size
if self.is_train:
crop_from_H = randint(0, frame_h - self.crop_size[1])
crop_from_W = randint(0, frame_w - self.crop_size[0])
else:
crop_from_H, crop_from_W = 0, 0
im1 = im1.crop((crop_from_W, crop_from_H, crop_from_W + self.crop_size[0], crop_from_H + self.crop_size[1]))
im2 = im2.crop((crop_from_W, crop_from_H, crop_from_W + self.crop_size[0], crop_from_H + self.crop_size[1]))
im3 = im3.crop((crop_from_W, crop_from_H, crop_from_W + self.crop_size[0], crop_from_H + self.crop_size[1]))
# random rotate.
if self.is_train and self.crop_size[0] == self.crop_size[1]:
angle = randint(0, 3)
im1 = im1.rotate(90 * angle)
im2 = im2.rotate(90 * angle)
im3 = im3.rotate(90 * angle)
im1 = to_tensor(im1)
im2 = to_tensor(im2)
im3 = to_tensor(im3)
return torch.stack([im1, im3], dim=1), im2, 0.5, name
# modified from:
# https://github.com/myungsub/CAIN/blob/master/data/snufilm.py
class SNU_FILM(data.Dataset):
def __init__(self, data_root='/dataset/SNU_FILM', mode='hard', crop=None):
super().__init__()
'''
:param data_root: ./data/SNU-FILM
:param mode: ['easy', 'medium', 'hard', 'extreme']
'''
self.data_root = data_root
test_fn = os.path.join(data_root, 'eval_modes', f'test-{mode}.txt')
with open(test_fn, 'r') as f:
self.frame_list = f.read().splitlines()
self.frame_list = [v.split(' ') for v in self.frame_list]
self.crop = crop if crop is None or isinstance(crop, tuple) else (crop, crop)
def __len__(self):
return len(self.frame_list)
def __getitem__(self, idx):
imgpaths = self.frame_list[idx]
# customization
# rename datapath to my path.
for i, imgpath in enumerate(imgpaths):
imgpaths[i] = imgpath.replace('data/SNU-FILM', self.data_root)
img1 = Image.open(os.path.join(self.data_root, imgpaths[0])).convert('RGB')
img2 = Image.open(os.path.join(self.data_root, imgpaths[1])).convert('RGB')
img3 = Image.open(os.path.join(self.data_root, imgpaths[2])).convert('RGB')
# customization
# cropping
if self.crop is not None:
img1 = img1.crop((0, 0, self.crop[1], self.crop[0]))
img2 = img1.crop((0, 0, self.crop[1], self.crop[0]))
img3 = img1.crop((0, 0, self.crop[1], self.crop[0]))
img1 = to_tensor(img1)
img2 = to_tensor(img2)
img3 = to_tensor(img3)
name = '/'.join(imgpaths[1].split('/')[-3:])
return torch.stack([img1, img3], dim=1), img2, 0.5, name
# referenced LDMVFI repo.
class Middlebury_others(data.Dataset):
def __init__(self, path):
super().__init__()
self.input_path = os.path.join(path, 'other-color-allframes/other-data')
self.gt_path = os.path.join(path, 'other-gt-interp')
self.seqs = sorted(glob(os.path.join(self.gt_path, '*')))
def __len__(self):
return len(self.seqs)
def __getitem__(self, idx):
video = self.seqs[idx]
seq_name = video.split('/')[-1]
gt = to_tensor(Image.open(os.path.join(video, 'frame10i11.png')).convert('RGB'))
img1 = to_tensor(Image.open(os.path.join(self.input_path, seq_name, 'frame10.png')).convert('RGB'))
img3 = to_tensor(Image.open(os.path.join(self.input_path, seq_name, 'frame11.png')).convert('RGB'))
return torch.stack([img1, img3], dim=1), gt, 0.5, seq_name
# modified from softsplat repo.
class Xiph(data.Dataset):
def __init__(self, path, setting='4K', extract_frames=True, from_link=False):
super().__init__()
self.path = os.path.join(path, 'frames')
self.setting = setting
os.makedirs(self.path, exist_ok=True)
if extract_frames:
if len(glob(f'{self.path}/BoxingPractice-*.png')) != 100:
if from_link:
os.system(f'ffmpeg -i https://media.xiph.org/video/derf/ElFuente/Netflix_BoxingPractice_4096x2160_60fps_10bit_420.y4m -pix_fmt rgb24 -vframes 100 {self.path}/BoxingPractice-%03d.png')
else:
os.system(f'ffmpeg -i {path}/Netflix_BoxingPractice_4096x2160_60fps_10bit_420.y4m -pix_fmt rgb24 -vframes 100 {self.path}/BoxingPractice-%03d.png')
# end
if len(glob(f'{self.path}/Crosswalk-*.png')) != 100:
if from_link:
os.system(f'ffmpeg -i https://media.xiph.org/video/derf/ElFuente/Netflix_Crosswalk_4096x2160_60fps_10bit_420.y4m -pix_fmt rgb24 -vframes 100 {self.path}/Crosswalk-%03d.png')
else:
os.system(f'ffmpeg -i {path}/Netflix_Crosswalk_4096x2160_60fps_10bit_420.y4m -pix_fmt rgb24 -vframes 100 {self.path}/Crosswalk-%03d.png')
# end
if len(glob(f'{self.path}/DrivingPOV-*.png')) != 100:
if from_link:
os.system(f'ffmpeg -i https://media.xiph.org/video/derf/Chimera/Netflix_DrivingPOV_4096x2160_60fps_10bit_420.y4m -pix_fmt rgb24 -vframes 100 {self.path}/DrivingPOV-%03d.png')
else:
os.system(f'ffmpeg -i {path}/Netflix_DrivingPOV_4096x2160_60fps_10bit_420.y4m -pix_fmt rgb24 -vframes 100 {self.path}/DrivingPOV-%03d.png')
# end
if len(glob(f'{self.path}/FoodMarket-*.png')) != 100:
if from_link:
os.system(f'ffmpeg -i https://media.xiph.org/video/derf/ElFuente/Netflix_FoodMarket_4096x2160_60fps_10bit_420.y4m -pix_fmt rgb24 -vframes 100 {self.path}/FoodMarket-%03d.png')
else:
os.system(f'ffmpeg -i {path}/Netflix_FoodMarket_4096x2160_60fps_10bit_420.y4m -pix_fmt rgb24 -vframes 100 {self.path}/FoodMarket-%03d.png')
# end
if len(glob(f'{self.path}/FoodMarket2-*.png')) != 100:
if from_link:
os.system(f'ffmpeg -i https://media.xiph.org/video/derf/ElFuente/Netflix_FoodMarket2_4096x2160_60fps_10bit_420.y4m -pix_fmt rgb24 -vframes 100 {self.path}/FoodMarket2-%03d.png')
else:
os.system(f'ffmpeg -i {path}/Netflix_FoodMarket2_4096x2160_60fps_10bit_420.y4m -pix_fmt rgb24 -vframes 100 {self.path}/FoodMarket2-%03d.png')
# end
if len(glob(f'{self.path}/RitualDance-*.png')) != 100:
if from_link:
os.system(f'ffmpeg -i https://media.xiph.org/video/derf/ElFuente/Netflix_RitualDance_4096x2160_60fps_10bit_420.y4m -pix_fmt rgb24 -vframes 100 {self.path}/RitualDance-%03d.png')
else:
os.system(f'ffmpeg -i {path}/Netflix_RitualDance_4096x2160_60fps_10bit_420.y4m -pix_fmt rgb24 -vframes 100 {self.path}/RitualDance-%03d.png')
# end
if len(glob(f'{self.path}/SquareAndTimelapse-*.png')) != 100:
if from_link:
os.system(f'ffmpeg -i https://media.xiph.org/video/derf/ElFuente/Netflix_SquareAndTimelapse_4096x2160_60fps_10bit_420.y4m -pix_fmt rgb24 -vframes 100 {self.path}/SquareAndTimelapse-%03d.png')
else:
os.system(f'ffmpeg -i {path}/Netflix_SquareAndTimelapse_4096x2160_60fps_10bit_420.y4m -pix_fmt rgb24 -vframes 100 {self.path}/SquareAndTimelapse-%03d.png')
# end
if len(glob(f'{self.path}/Tango-*.png')) != 100:
if from_link:
os.system(f'ffmpeg -i https://media.xiph.org/video/derf/ElFuente/Netflix_Tango_4096x2160_60fps_10bit_420.y4m -pix_fmt rgb24 -vframes 100 {self.path}/Tango-%03d.png')
else:
os.system(f'ffmpeg -i {path}/Netflix_Tango_4096x2160_60fps_10bit_420.y4m -pix_fmt rgb24 -vframes 100 {self.path}/Tango-%03d.png')
# end
# make triplets
self.triplets = []
for strFile in ['BoxingPractice', 'Crosswalk', 'DrivingPOV', 'FoodMarket', 'FoodMarket2', 'RitualDance', 'SquareAndTimelapse', 'Tango']:
for intFrame in range(2, 99, 2):
npyOne = f'{self.path}/' + strFile + '-' + str(intFrame - 1).zfill(3) + '.png'
npyTwo = f'{self.path}/' + strFile + '-' + str(intFrame + 1).zfill(3) + '.png'
npyTruth = f'{self.path}/' + strFile + '-' + str(intFrame).zfill(3) + '.png'
triplet = (npyOne, npyTruth, npyTwo)
self.triplets.append(triplet)
def __len__(self):
return len(self.triplets)
def __getitem__(self, idx):
triplet = self.triplets[idx]
npyOne = cv2.imread(triplet[0], flags=-1)
npyTruth = cv2.imread(triplet[1], flags=-1)
npyTwo = cv2.imread(triplet[2], flags=-1)
if self.setting == '2K': # resized
npyOne = cv2.resize(src=npyOne, dsize=(2048, 1080), fx=0.0, fy=0.0, interpolation=cv2.INTER_AREA)
npyTwo = cv2.resize(src=npyTwo, dsize=(2048, 1080), fx=0.0, fy=0.0, interpolation=cv2.INTER_AREA)
npyTruth = cv2.resize(src=npyTruth, dsize=(2048, 1080), fx=0.0, fy=0.0, interpolation=cv2.INTER_AREA)
elif self.setting == '4K': # cropped
npyOne = npyOne[540:-540, 1024:-1024, :]
npyTwo = npyTwo[540:-540, 1024:-1024, :]
npyTruth = npyTruth[540:-540, 1024:-1024, :]
npyOne = cv2.cvtColor(npyOne, cv2.COLOR_BGR2RGB)
npyTruth = cv2.cvtColor(npyTruth, cv2.COLOR_BGR2RGB)
npyTwo = cv2.cvtColor(npyTwo, cv2.COLOR_BGR2RGB)
tenOne = torch.FloatTensor(np.ascontiguousarray(npyOne.transpose(2, 0, 1).astype(np.float32) * (1.0 / 255.0)))
tenTwo = torch.FloatTensor(np.ascontiguousarray(npyTwo.transpose(2, 0, 1).astype(np.float32) * (1.0 / 255.0)))
tenGT = torch.FloatTensor(np.ascontiguousarray(npyTruth.transpose(2, 0, 1).astype(np.float32) * (1.0 / 255.0)))
name = '/'.join(triplet[1].split('/')[-2:])
return torch.stack([tenOne, tenTwo], dim=1), tenGT, 0.5, name
# for demo, taking video inputs
class VideoData(data.Dataset):
def __init__(self, video, save_as_png=False):
super().__init__()
buffer_path = None
if save_as_png:
buffer_path = f'./frames_buffer'
self.frames_list = video2frames(video, output_path=buffer_path)
self.save_as_png = save_as_png
def __len__(self):
return len(self.frames_list) - 1
def __getitem__(self, idx):
frame0 = self.frames_list[idx]
frame1 = self.frames_list[idx + 1]
if self.save_as_png:
frame0 = cv2.imread(frame0, cv2.IMREAD_COLOR)
frame1 = cv2.imread(frame1, cv2.IMREAD_COLOR)
frame0 = to_tensor(cv2.cvtColor(frame0, cv2.COLOR_BGR2RGB))
frame1 = to_tensor(cv2.cvtColor(frame1, cv2.COLOR_BGR2RGB))
return frame0, frame1