-
Notifications
You must be signed in to change notification settings - Fork 2
/
utils.py
133 lines (115 loc) · 3.82 KB
/
utils.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
import os
from time import time
import datetime
import numpy as np
from skimage import io
import cv2
import yaml
from functools import wraps
import requests
class Progress:
"""Progress bar"""
def __init__(self, i_list):
self.__list = list(i_list)
self.total = len(self.__list)
self.current = -1
self.__bar_length = 0
self.__begin_time = time()
self.__start = 0
self.__recently = []
def __iter__(self):
return self
def __next__(self):
self.__update()
if self.current < self.total:
return self.__list[self.current]
raise StopIteration
def __update_bar_length(self, bar):
terminal_lenght = os.get_terminal_size()[0]
external_bar_length = int(len(bar)-self.__bar_length)
self.__bar_length = terminal_lenght-external_bar_length-2
def __per_sec(self):
prev = time()
ps = (1 / (prev - self.__start)) if self.__start != 0 else 0
self.__recently.pop(0) if len(self.__recently) >= 100 else None
self.__recently.append(ps) if ps <= 100 else None
self.__start = prev
return np.mean(self.__recently)
def __time_left(self):
ps = self.__per_sec()
sec_left = round((self.total - self.current) / ps) if ps != 0 else 0
time_format = datetime.timedelta(seconds=sec_left)
return time_format
def __time_total(self):
sec_total = round(time()-self.__begin_time)
time_format = datetime.timedelta(seconds=sec_total)
return time_format
def __bar(self):
percent = self.current / self.total
completed = int(percent * self.__bar_length) * '█'
padding = int(self.__bar_length - len(completed)) * '.'
return f'Progress: {self.current}/{self.total} |{completed}{padding}| {int(percent*100)}% in {self.__time_left()}s > {self.__time_total()}s '
def __update(self):
"""Finish current state and move to next state"""
self.current += 1
bar = self.__bar()
ending = '\n' if self.current == self.total else '\r'
self.__update_bar_length(bar)
print(bar, end=ending)
def crop_background(image, grayscale=False):
"""Crop black background only"""
if not type(image).__module__ == np.__name__:
img_arr = io.imread(image, True)
else:
img_arr = image
gray = img_arr[:,:,0] if img_arr.ndim > 2 else img_arr
_, thresholded = cv2.threshold(gray, 0, 255, cv2.THRESH_OTSU)
x, y, w, h = cv2.boundingRect(thresholded)
output = (gray if grayscale else image)[y:y+h, x:x+w]
return output
def load_config(config_name, args=None):
"""Load config file"""
with open('config.yaml') as f:
config = yaml.full_load(f)[config_name]
if args is not None:
for arg, value in args.__dict__.items():
if value is not None:
config[arg] = value
return config
def measure(func):
"""Measure the runtime"""
@wraps(func)
def _time(*args, **kwargs):
start = time()
try:
return func(*args, **kwargs)
finally:
end_ = time() - start
print(f"Done {func.__name__} in {round(end_, 2)}s")
return _time
def download_weight(model_name, url=False):
models = {
'rotate_180.pkl': "1laOUDHBEOtazXM20x2DCXIZ7zBH4uWWB",
'craft_mlt_25k.pth': "1fBUOVdtv4r6UM4rOTjaZpamcs3UIwzj_",
'craft_refiner_CTW1500.pth': "1QC0hXbyNX-yW69g42RR0ZB74L6jqdRDg",
}
if not url:
url = f"https://drive.google.com/uc?export=download&id={models[model_name]}"
else:
url = model_name
model_name = model_name.split('/')[-1]
os.mkdir('weights') if not os.path.exists('weights') else None
weight_path = f'weights/{model_name}'
if os.path.exists(weight_path):
return weight_path
print(f"Weight '{model_name}' not found, requesting...", end='\r')
with requests.get(url, stream=True) as r:
r.raise_for_status()
weight_folder = 'weights'
os.mkdir(weight_folder) if not os.path.exists(weight_folder) else None
weight_path = f'{weight_folder}/{model_name}'
with open(weight_path, 'wb') as f:
print(f"Download weight into '{weight_path}'"+' '*10)
for chunk in Progress(r.iter_content(chunk_size=1000)):
f.write(chunk)
return weight_path