-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathdata.py
158 lines (126 loc) · 6.04 KB
/
data.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
import os
import math
from os.path import join
import torch
import numpy as np
from PIL import Image, ImageDraw
from torch.utils.data import IterableDataset, Dataset
from torchvision.transforms import ToTensor, Normalize, Compose
def get_mean_std(pil_image):
np_img = np.array(pil_image) / 255.0
std = np.std(np_img, axis=(0,1))
return np.mean(np_img, axis=(0,1)), std if np.linalg.norm(std) > 1e-4 else np.ones(std.shape)
class SingleImageTrainDataset(IterableDataset):
def __init__(self, image_path):
super().__init__()
self.img = Image.open(join(image_path, "diffuse.png")).convert("RGB")
self.normal = Image.open(join(image_path, "normal.png")).convert("RGB")
self.roughness = Image.open(join(image_path, "roughness.png")).convert("RGB")
self.specular = Image.open(join(image_path, "specular.png")).convert("RGB")
self.mask = Image.open(join(image_path, "mask.png"))
self.transforms_mean_std = {
"normal": Compose([ToTensor(), Normalize(*get_mean_std(self.normal))]),
"diffuse": Compose([ToTensor(), Normalize(*get_mean_std(self.img))]),
"roughness": Compose([ToTensor(), Normalize(*get_mean_std(self.roughness))]),
"specular": Compose([ToTensor(), Normalize(*get_mean_std(self.specular))])
}
self.transforms = Compose([
ToTensor(),
Normalize(0.5, 0.5) # Change range [0,1] -> [-1, 1]
])
self.mask_transform = Compose([
ToTensor()
])
def __iter__(self):
def next_crop():
# Draw new coordinates for our image until not overlapping the center crop (256x256)
w, h = self.img.size
while True:
x, y = np.random.randint(0, w - 256), np.random.randint(0, h - 256)
box = (x, y, x + 256, y + 256)
yield torch.cat([self.transforms_mean_std["normal"](self.normal.crop(box)),
self.transforms_mean_std["diffuse"](self.img.crop(box)),
self.transforms_mean_std["roughness"](self.roughness.crop(box)),
self.transforms_mean_std["specular"](self.specular.crop(box))], dim=0), self.mask_transform(self.mask.crop(box))
return iter(next_crop())
class SingleImageTestDataset(Dataset):
def __init__(self, image_path):
super().__init__()
self.img = Image.open(join(image_path, "diffuse.png")).convert("RGB")
self.normal = Image.open(join(image_path, "normal.png")).convert("RGB")
self.roughness = Image.open(join(image_path, "roughness.png")).convert("RGB")
self.specular = Image.open(join(image_path, "specular.png")).convert("RGB")
self.mask = Image.open(join(image_path, "mask.png"))
self.transforms = Compose([
ToTensor(),
Normalize(0.5, 0.5) # Change range [0,1] -> [-1, 1]
])
self.mask_transform = Compose([
ToTensor()
])
self.transforms_mean_std = {
"normal": Compose([ToTensor(), Normalize(*get_mean_std(self.normal))]),
"diffuse": Compose([ToTensor(), Normalize(*get_mean_std(self.img))]),
"roughness": Compose([ToTensor(), Normalize(*get_mean_std(self.roughness))]),
"specular": Compose([ToTensor(), Normalize(*get_mean_std(self.specular))])
}
print(self.transforms_mean_std)
def __len__(self):
return 1
def __getitem__(self, idx):
return torch.cat([self.transforms_mean_std["normal"](self.normal),
self.transforms_mean_std["diffuse"](self.img),
self.transforms_mean_std["roughness"](self.roughness),
self.transforms_mean_std["specular"](self.specular)], dim=0), self.mask_transform(self.mask)
class MaskDataset(IterableDataset):
def __init__(self):
super().__init__()
def __iter__(self):
return iter(mask_generator(256, 256))
def mask_generator(H, W):
min_num_vertex = 4
max_num_vertex = 12
mean_angle = 2*math.pi / 5
angle_range = 2*math.pi / 15
min_width = 12
max_width = 40
while True:
# Code from ContextualAttention
average_radius = math.sqrt(H*H+W*W) / 8
mask = Image.new('L', (W, H), 0)
for _ in range(np.random.randint(1, 4)):
num_vertex = np.random.randint(min_num_vertex, max_num_vertex)
angle_min = mean_angle - np.random.uniform(0, angle_range)
angle_max = mean_angle + np.random.uniform(0, angle_range)
angles = []
vertex = []
for i in range(num_vertex):
if i % 2 == 0:
angles.append(2*math.pi - np.random.uniform(angle_min, angle_max))
else:
angles.append(np.random.uniform(angle_min, angle_max))
h, w = mask.size
vertex.append((int(np.random.randint(0, w)), int(np.random.randint(0, h))))
for i in range(num_vertex):
r = np.clip(
np.random.normal(loc=average_radius, scale=average_radius//2),
0, 2*average_radius)
new_x = np.clip(vertex[-1][0] + r * math.cos(angles[i]), 0, w)
new_y = np.clip(vertex[-1][1] + r * math.sin(angles[i]), 0, h)
vertex.append((int(new_x), int(new_y)))
draw = ImageDraw.Draw(mask)
width = int(np.random.uniform(min_width, max_width))
draw.line(vertex, fill=1, width=width)
for v in vertex:
draw.ellipse((v[0] - width//2,
v[1] - width//2,
v[0] + width//2,
v[1] + width//2),
fill=1)
if np.random.normal() > 0:
mask.transpose(Image.FLIP_LEFT_RIGHT)
if np.random.normal() > 0:
mask.transpose(Image.FLIP_TOP_BOTTOM)
mask = np.asarray(mask, np.float32)
mask = np.reshape(mask, (1, H, W))
yield mask