Question in the normalization.py code

[eastchun]

In normalization.py,

1. You resize/normalize/crop each of source label image according to the size ratio between the 2nd source label image and 2nd target label image (i.e., 00001.png). But, when you calculate the head position of the normalized label image in Line 77, you used the coordinate of the 1st entry (i.e., head position of the 1st original source label image):

   Line 77: source_cord_y, source_cord_x = pose_cord[0]

Shouldn't the above be changed to the 2nd entry (i.e., pose_cord[0] -> pose_cord[1]) as follow:
source_cord_y, source_cord_x = pose_cord[1] ???

2. In line 84, when you crop the head image of the normalized label image, you used the old label image:

   Line 84: head = img[int(new_head_x - crop_size): int(new_head_x + crop_size),

Shouldn't the above be changed to crop the 'new_source' instead of 'img' as follow:
head = new_source[int(new_head_x - crop_size): int(new_head_x + crop_size), ???