import numpy as np
from scipy import signal
import pickle
"""
This file is to prepare the pairwise distribution for fully connected model
"""
print(pickle.format_version)
y_train = np.load('y_train_flic.npy')
print(y_train.shape)
coefs = np.array([[1, 8, 28, 56, 70, 56, 28, 8, 1]], dtype=np.uint16) / 256
kernel = coefs.T @ coefs
joint_ids = ['lsho', 'lelb', 'lwri', 'rsho', 'relb', 'rwri', 'lhip', 'rhip', 'nose', 'torso']
joint_dependece = {'lsho': ['nose', 'lelb'], 'lelb': ['lsho', 'lwri'], 'lwri': ['lelb'],
'rsho': ['nose', 'relb'], 'relb': ['rsho', 'rwri'], 'rwri': ['relb'],
'lhip': ['nose'], 'rhip': ['nose'], 'nose': ['lsho', 'rsho', 'lhip', 'rhip']}
dict = {'lsho': 0, 'lelb': 1, 'lwri': 2, 'rsho': 3, 'relb': 4, 'rwri': 5, 'lhip': 6,
'lkne': 7, 'lank': 8, 'rhip': 9, 'rkne': 10, 'rank': 11, 'leye': 12, 'reye': 13,
'lear': 14, 'rear': 15, 'nose': 16, 'msho': 17, 'mhip': 18, 'mear': 19, 'mtorso': 20,
'mluarm': 21, 'mruarm': 22, 'mllarm': 23, 'mrlarm': 24, 'mluleg': 25, 'mruleg': 26,
'mllleg': 27, 'mrlleg': 28}
def compute_pairwise_distribution(joint, cond_j):
"""
This computes a single histogram for a pair of (joint, cond_joint), and applies gaussian smoothing
:param joint: e.g. 'lsho'
:param cond_j: e.g. 'nose'
:return: 120 x 180 pairwise distribution
"""
hp_height = y_train.shape[1]
hp_width = y_train.shape[2]
pd = np.zeros([hp_height * 2, hp_width * 2]) # the pairwise distribution is twice the size of the heat map
# print(pd.shape)
for i in range(y_train.shape[0]): # for every single image, we note the distance between the joint and cond_j
img_j = y_train[i, :, :, joint_ids.index(joint)]
img_cj = y_train[i, :, :, joint_ids.index(cond_j)]
xj, yj = np.where(img_j == np.max(img_j))
xcj, ycj = np.where(img_cj == np.max(img_cj))
pd[hp_height + (xj - xcj), hp_width + (yj - ycj)] += 1 # count for the histgram
pd = pd / np.float32(np.sum(pd))
pd = signal.convolve2d(pd, kernel, mode='same', boundary='fill', fillvalue=0)
return pd
pairwise_distribution = {}
for joint in joint_ids:
for cond_j in joint_ids:
if cond_j is not joint:
print(joint + '_' + cond_j)
pairwise_distribution[joint + '_' + cond_j] = compute_pairwise_distribution(joint, cond_j)
# print(type(pairwise_distribution['rhip_nose']),pairwise_distribution['rhip_nose'].shape, np.sum(pairwise_distribution['rhip_nose']))
# import matplotlib.pyplot as plt
# plt.figure(1)
# plt.imshow(np.uint8(pairwise_distribution['lhip_nose'])*255)
# plt.figure(2)
# plt.imshow(np.uint8(pairwise_distribution['rhip_nose'])*255)
# plt.show()
print('number of pairs is: ', len(pairwise_distribution))
with open('pairwise_distribution.pickle', 'wb') as handle:
pickle.dump(pairwise_distribution, handle, protocol=pickle.HIGHEST_PROTOCOL)
