Datasets:
jblitzar
/
github-python-corpus

Dataset card Data Studio Files
xet
Community
text
stringlengths
1
93.6k
output_matrix = compute_output_matrix(gt, prediction, output_matrix)
total_num += label.shape[0]
if (step+1) % config['skip_step'] == 0:
print '%s %s] %d. iou updating' \
% (str(datetime.datetime.now()), str(os.getpid()), total_num)
print 'mIoU: ', compute_iou(output_matrix)
step += 1
except tf.errors.OutOfRangeError:
print 'mIoU: ', compute_iou(output_matrix), 'total_data: ', total_num
break
def main():
args = PARSER.parse_args()
if args.config:
file_address = open(args.config)
config = yaml.load(file_address)
else:
print '--config config_file_address missing'
test_func(config)
if __name__ == '__main__':
main()
# <FILESEP>
import cv2, numpy, random
import os
from PIL import Image, ImageDraw
#Prepare the hats for use later
hats = []
hats_portion = []
for i in range(1,8):
hat = Image.open('./xmas/' + 'hat' + str(i) + '.png')
hats.append(hat)
hats_portion.append(hat.size[0]/hat.size[1])
#load photo frames
frame_w = Image.open('./xmas/frame_w.png')
frame_h = Image.open('./xmas/frame_h.png')
frame_s = Image.open('./xmas/frame_s.png')
ok_msg = '你的圣诞节靓照已经生成好啦，请访问以下链接查看，长按图片可保存： '
auto_reply = '嗨！欢迎来玩儿，活动正在进行中！你可以在这里向本公众号发送现场拍摄的照片，我们会自动为照片中的人物带上圣诞帽或鹿角，生成有浓浓圣诞节氛围的照片～「本消息为自动发送」'
def gen_pic(MediaId):
def get_hat(x,y,w,h):
#set the probability of every hat
num = random.randint(1,100)
if num in range(1,17):
#hat1
(hat_num, offset1, offset2, offset3) = (0, 1.2, .05, .67)
elif num in range(17,33):
#hat2
(hat_num, offset1, offset2, offset3) = (1, 1.3, -.4, .62)
elif num in range(33,49):
#hat3
(hat_num, offset1, offset2, offset3) = (2, .9, .05, .8)
elif num in range(91,101):
#green hat
(hat_num, offset1, offset2, offset3) = (3, 1.2, .05, .67)
elif num in range(49,65):
#jiao1
(hat_num, offset1, offset2, offset3) = (4, 1.2, -.1, 1.2)
elif num in range(65,81):
#jiao2
(hat_num, offset1, offset2, offset3) = (5, 1, 0, 1.2)
elif num in range(81,91):
#tree
(hat_num, offset1, offset2, offset3) = (6, .9, .05, 1)
hat_portion = hats_portion[hat_num]
(hat_w, hat_h) = (int(w*offset1), int(w*offset1/hat_portion))
print('hat size:',hat_w,hat_h)
hatter = hats[hat_num].resize((hat_w, hat_h))
(hat_x, hat_y) = (int(x+w*offset2), int(y-hat_h*offset3))
hat_pos = (hat_x, hat_y)
print('hat at:',hat_x,hat_y)
return (hatter, hat_pos)
def std_size(imagePath):
pic = Image.open(imagePath)
portion = pic.size[0]/pic.size[1]
if portion < 1:
#portrait
if portion <= 0.75:
pic_w = 960
pic_h = round(960/portion)
box = (0,round((pic_h-1280)/2),960,round(pic_h/2+640))
if portion > 0.75:
pic_h = 1280
pic_w = round(1280*portion)
box = (round((pic_w-960)/2),0,round(pic_w/2+480),1280)