-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathresults_summary.py
executable file
·306 lines (239 loc) · 13.2 KB
/
results_summary.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
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
#!/usr/bin/env python
import rospy
import roslib
roslib.load_manifest('multimodal_pose')
from std_msgs.msg import String
from sensor_msgs.msg import CompressedImage
from sensor_msgs.msg import Image
from hrl_msgs.msg import FloatArrayBare
import numpy as np
import cv2
from cv_bridge import CvBridge, CvBridgeError
from ar_track_alvar_msgs.msg import AlvarMarkers
import os.path as osp
from camera import Camera
import pickle
import time
import imutils
import math
import cPickle as pkl
import os
SHORT = False
import rospy
# from hrl_lib.util import load_pickle
def load_pickle(filename):
with open(filename, 'rb') as f:
return pickle.load(f)
def get_heightweight_from_betas(betas):
height, weight = 0, 0
return height, weight
if __name__ == '__main__':
RESULT_TYPE = "synth"
if RESULT_TYPE == "real":
participant_list = ["S103",
"S104",
"S107",
"S114",
"S118",
"S121",
"S130",
"S134",
"S140",
"S141",
"S145",
"S151",
"S163",
"S165", #at least 3 pc corrupted
"S170",
"S179",
"S184",
"S187",
"S188", #1 bad prone posture classified as supine, 2 pc corrupted
"S196",]
#participant_list=["S188"]
general = [[]]
general_plo = [[]]
general_supine = [[]]
general_plo_supine = [[]]
hands_behind_head = [[]]
prone_hands_up = [[]]
crossed_legs = [[]]
straight_limbs = [[]]
#NETWORK_2 = "1.0rtojtdpth_angleadj_tnhFIXN_htwt_calnoise"
NETWORK_2 = "0.5rtojtdpth_depthestin_angleadj_tnhFIXN_htwt_calnoise"
#NETWORK_2 = "NONE-200e"
#NETWORK_2 = "BASELINE"
POSE_TYPE = "2"
DATA_QUANT = "46K"
recall_list = []
precision_list = []
overlap_d_err_list = []
v_limb_to_gt_err_list = []
v_to_gt_err_list = []
gt_to_v_err_list = []
for participant in participant_list:
participant_directory = "/media/henry/multimodal_data_2/CVPR2020_study/"+participant
participant_info = load_pickle(participant_directory + "/participant_info.p")
pose_type_list = participant_info['pose_type']
#if participant_info['gender'] == 'f': continue
print "participant directory: ", participant_directory
if True:#DATA_QUANT == "46K":
current_results_dict = load_pickle("/media/henry/multimodal_data_2/data/final_results/"+DATA_QUANT+"_"
+NETWORK_2+"_V2/results_real_"+DATA_QUANT+"_"
+participant+"_"+POSE_TYPE+"_"+NETWORK_2+".p")
else:
current_results_dict = load_pickle("/media/henry/multimodal_data_2/data/final_results/"+NETWORK_2+"_V2/results_real_"
+participant+"_"+POSE_TYPE+"_"+NETWORK_2+".p")
print "/media/henry/multimodal_data_2/data/final_results/"+NETWORK_2+"/results_real_"+participant+"_"+POSE_TYPE+"_"+NETWORK_2+".p"
#for entry in current_results_dict:
# print entry
#precision =
#print participant
#print len(current_results_dict['recall'])
print len(current_results_dict['recall'])
#to test posture
idx_num = -1
if POSE_TYPE == "1":
num_ims = 5
elif POSE_TYPE == "2":
num_ims = 48
recall_list_curr = []
precision_list_curr = []
overlap_d_err_list_curr = []
v_limb_to_gt_err_list_curr = []
v_to_gt_err_list_curr = []
gt_to_v_err_list_curr = []
for i in range(num_ims):
partition_type = pose_type_list[i]
print partition_type
if participant == "S114" and POSE_TYPE == "2" and i in [26, 29]:
print "skipping", i, partition_type
continue #these don't have point clouds
elif participant == "S165" and POSE_TYPE == "2" and i in [1, 3, 15]:
print "skipping", i, partition_type
continue #these don't have point clouds
elif participant == "S188" and POSE_TYPE == "2" and i in [5, 17, 21]:
print "skipping", i, partition_type
continue
elif participant == "S145" and POSE_TYPE == "1" and i in [0]:
print "skipping", i
continue
else: idx_num += 1
#body_roll_rad = current_results_dict['body_roll_rad'][idx_num]
#if partition_type in ['phu']:
recall_list_curr.append(current_results_dict['recall'][idx_num])
precision_list_curr.append(current_results_dict['precision'][idx_num])
overlap_d_err_list_curr.append( current_results_dict['overlap_d_err'][idx_num])
v_limb_to_gt_err_list_curr.append(current_results_dict['v_limb_to_gt_err'][idx_num])
v_to_gt_err_list_curr.append(current_results_dict['v_to_gt_err'][idx_num])
gt_to_v_err_list_curr.append(current_results_dict['gt_to_v_err'][idx_num])
'''
if idx_num in [3] and participant not in ["S145", "S188", "S140"]:
#if partition_type in ['supine_plo', 'rollpi_plo']:
recall_list.append(current_results_dict['recall'][idx_num])
precision_list.append(current_results_dict['precision'][idx_num])
overlap_d_err_list.append( current_results_dict['overlap_d_err'][idx_num])
v_limb_to_gt_err_list.append(current_results_dict['v_limb_to_gt_err'][idx_num])
v_to_gt_err_list.append(current_results_dict['v_to_gt_err'][idx_num])
gt_to_v_err_list.append(current_results_dict['gt_to_v_err'][idx_num]) #for 140 get supine from last
elif idx_num in [1] and participant in ["S188"]:
recall_list.append(current_results_dict['recall'][idx_num])
precision_list.append(current_results_dict['precision'][idx_num])
overlap_d_err_list.append( current_results_dict['overlap_d_err'][idx_num])
v_limb_to_gt_err_list.append(current_results_dict['v_limb_to_gt_err'][idx_num])
v_to_gt_err_list.append(current_results_dict['v_to_gt_err'][idx_num])
gt_to_v_err_list.append(current_results_dict['gt_to_v_err'][idx_num]) #for 140 get supine from last
elif idx_num in [2] and participant in ["S145"]:
recall_list.append(current_results_dict['recall'][idx_num])
precision_list.append(current_results_dict['precision'][idx_num])
overlap_d_err_list.append( current_results_dict['overlap_d_err'][idx_num])
v_limb_to_gt_err_list.append(current_results_dict['v_limb_to_gt_err'][idx_num])
v_to_gt_err_list.append(current_results_dict['v_to_gt_err'][idx_num])
gt_to_v_err_list.append(current_results_dict['gt_to_v_err'][idx_num]) #for 140 get supine from last
elif idx_num in [2] and participant in ["S140"]:
recall_list.append(current_results_dict['recall'][idx_num])
precision_list.append(current_results_dict['precision'][idx_num])
overlap_d_err_list.append( current_results_dict['overlap_d_err'][idx_num])
v_limb_to_gt_err_list.append(current_results_dict['v_limb_to_gt_err'][idx_num])
v_to_gt_err_list.append(current_results_dict['v_to_gt_err'][idx_num])
gt_to_v_err_list.append(current_results_dict['gt_to_v_err'][idx_num]) #for 140 get supine from last'''
recall_list.append(np.mean(recall_list_curr))
precision_list.append(np.mean(precision_list_curr))
overlap_d_err_list.append(np.mean(overlap_d_err_list_curr))
v_limb_to_gt_err_list.append(np.mean(v_limb_to_gt_err_list_curr))
v_to_gt_err_list.append(np.mean(v_to_gt_err_list_curr))
gt_to_v_err_list.append(np.mean(gt_to_v_err_list_curr))
#print curr_gt_to_v_err
print len(v_to_gt_err_list), 'ct list'
#break
#height, weight = get_heightweight_from_betas(current_results_dict['betas'])
#print "average precision: ", np.mean(precision_list)
# print "average recall: ", np.mean(recall_list)
#print "average overlap depth err: ", np.mean(overlap_d_err_list)
print "average v to gt err: ", np.mean(v_to_gt_err_list)*100
print "average gt to v err: ", np.mean(gt_to_v_err_list)*100
print "mean 3D err: ", np.mean([np.mean(v_to_gt_err_list), np.mean(gt_to_v_err_list)])
elif RESULT_TYPE == "synth":
#NETWORK_1 = "1.0rtojtdpth_depthestin_angleadj_tnhFIXN_htwt_calnoise"
NETWORK_2 = "0.5rtojtdpth_depthestin_angleadj_tnhFIXN"
#NETWORK_2 = "1.0rtojtdpth_angleadj_tnhFIXN_calnoise"
#NETWORK_2 = "NONE-200e"
#NETWORK_2 = "BASELINE"
DATA_QUANT = "184K"
filename_list = ["test_rollpi_f_lay_set23to24_1500_",
"test_rollpi_m_lay_set23to24_1500_",
"test_rollpi_plo_f_lay_set23to24_1500_",
"test_rollpi_plo_m_lay_set23to24_1500_",
"test_roll0_f_lay_set14_500_",
"test_roll0_m_lay_set14_500_",
"test_roll0_plo_f_lay_set14_500_",
"test_roll0_plo_m_lay_set14_500_",
"test_roll0_plo_hbh_f_lay_set4_500_",
"test_roll0_plo_hbh_m_lay_set1_500_",
"test_roll0_plo_phu_f_lay_set1pa3_500_",
"test_roll0_plo_phu_m_lay_set1pa3_500_",
"test_roll0_sl_f_lay_set1both_500_",
"test_roll0_sl_m_lay_set1both_500_",
"test_roll0_xl_f_lay_set1both_500_",
"test_roll0_xl_m_lay_set1both_500_"
]
import math
recall_avg_list = []
precision_avg_list = []
overlap_d_err_avg_list = []
v_to_gt_err_avg_list = []
gt_to_v_err_avg_list = []
joint_err_list = []
for filename in filename_list:
#current_results_dict = load_pickle("/media/henry/multimodal_data_1/data/final_results/"+DATA_QUANT+"_"
# +NETWORK_2+"/results_synth_"+DATA_QUANT+"_"+filename+NETWORK_2+".p")
#current_results_dict = load_pickle("/media/henry/multimodal_data_2/data/final_results/"+NETWORK_2+"/results_synth_"+filename+NETWORK_2+".p")
#current_results_dict = load_pickle("/home/henry/data/final_results/"+NETWORK_2+"/results_synth_"+filename+NETWORK_2+".p")
current_results_dict = load_pickle("/media/henry/multimodal_data_2/data/final_results/"+DATA_QUANT+"_"+NETWORK_2+"/results_synth_"+DATA_QUANT+"_"+filename+NETWORK_2+".p")
for entry in current_results_dict:
print entry
#print current_results_dict['j_err'], 'j err'
#precision =
for i in range(len(current_results_dict['v_to_gt_err'])):
#recall_avg_list.append(current_results_dict['recall'][i])
#if math.isnan(float(current_results_dict['precision'][i])): print "nan precision"
#else: precision_avg_list.append(current_results_dict['precision'][i])
#if math.isnan(float(current_results_dict['overlap_d_err'][i])): print "nan"
#else: overlap_d_err_avg_list.append(current_results_dict['overlap_d_err'][i])
v_to_gt_err_avg_list.append(current_results_dict['v_to_gt_err'][i])
gt_to_v_err_avg_list.append(current_results_dict['gt_to_v_err'][i])
#print curr_gt_to_v_err
joint_err_list.append(current_results_dict['j_err'][i])
#print np.shape(joint_err_list)
# break
#print np.min(overlap_d_err_avg_list), np.max(overlap_d_err_list)
print np.shape(joint_err_list)
print
print len(v_to_gt_err_avg_list)
#print "average precision: ", np.mean(precision_avg_list)
#print "average recall: ", np.mean(recall_avg_list)
#print "average overlap depth err: ", np.mean(overlap_d_err_avg_list)
print "average v to gt err: ", np.mean(v_to_gt_err_avg_list)*100
print "average gt to v err: ", np.mean(gt_to_v_err_avg_list)*100
print "mean 3D err: ", np.mean([np.mean(v_to_gt_err_avg_list), np.mean(gt_to_v_err_avg_list)])*100
print "mean joint err: ", np.mean(joint_err_list)*100