-
Notifications
You must be signed in to change notification settings - Fork 25
/
Copy pathanalyze.py
226 lines (184 loc) · 8.08 KB
/
analyze.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
import numpy as np
import matplotlib.pyplot as plt
from matplotlib.ticker import FormatStrFormatter
from matplotlib.colors import LogNorm
from scipy import stats
import os
import pickle
from utils import ROIs, count_violation_pairs, decode_data, cal_min_dists_all_frame, custom_simple_linear_regression
dict_dataset_names = {
'oxford_town': 'Oxford Town Center Dataset',
'mall': 'Mall Dataset',
'grand_central': 'Train Station Dataset'
}
path_results = 'results/test_faster_rcnn' # faster RCNN results
# path_results = 'additional_detectors/yolo_v4_darknet_official/results/result_yolo_v4_darknet_official'
def analyze_statistics(dataset):
print('=======================')
print('Processing %s ...' % dataset)
path_result = os.path.join(path_results, dataset)
path_analysis = os.path.join(path_results, 'analysis')
os.makedirs(path_result, exist_ok=True)
os.makedirs(path_analysis, exist_ok=True)
data = pickle.load(open(os.path.join(path_result, 'statistic_data.p'), 'rb'))
roi = ROIs[dataset]
x_min, x_max, y_min, y_max = roi
area = (x_max - x_min) * (y_max - y_min)
indexs_frame, ts_inference, pts_roi_all_frame, density, nums_ped = decode_data(data=data, roi=roi)
print('Mean inference time = %.6f' % np.mean(ts_inference))
all_min_dists, min_min_dists, avg_min_dists = cal_min_dists_all_frame(pts_roi_all_frame)
violations = count_violation_pairs(pts_all_frames=pts_roi_all_frame)
none_indexes = np.where(avg_min_dists == None)[0]
indexs_frame = np.delete(indexs_frame, none_indexes, 0)
density = np.delete(density, none_indexes, 0)
min_min_dists = np.delete(min_min_dists, none_indexes, 0)
avg_min_dists = np.delete(avg_min_dists, none_indexes, 0)
nums_ped = np.delete(nums_ped, none_indexes, 0)
violations = np.delete(violations, none_indexes, 0)
# violations = violations / nums_ped
# violations, density = density, violations
# figure 1 - min dists
fig = plt.figure(figsize=(5., 3.))
fig.subplots_adjust(left=0.12, right=0.97, top=0.9, bottom=0.15)
fig.suptitle(dict_dataset_names[dataset])
ax = fig.add_subplot(111)
ax.hist(all_min_dists, bins=100)
ax.grid()
ax.set_xlabel(r'Closest Distance ($m$)')
ax.set_ylabel('Count')
ax.set(xlim=(0, 10))
fig.savefig(os.path.join(path_analysis, dataset + '_closest_dists.png'))
plt.close(fig)
# figure 2 - min-min dists
fig = plt.figure(figsize=(5., 3.))
fig.subplots_adjust(left=0.12, right=0.97, top=0.9, bottom=0.15)
fig.suptitle(dict_dataset_names[dataset])
ax = fig.add_subplot(111)
ax.hist(min_min_dists, bins=100)
ax.grid()
ax.set_xlabel(r'Minimal Closest Distance ($m$)')
ax.set_ylabel('Count')
fig.savefig(os.path.join(path_analysis, dataset + '_min_closest_dists.png'))
plt.close(fig)
# figure 3 - density, min dist, min_min dist over time
t_max = 300
if dataset == 'oxford_town':
ts = indexs_frame / 10
elif dataset == 'mall':
ts = indexs_frame / 1
elif dataset == 'grand_central':
ts = indexs_frame / 25
else:
raise Exception('invalid dataset')
extracted = ts <= t_max
fig = plt.figure(figsize=(11.47, 4.33))
fig.subplots_adjust(left=0.07, bottom=0.12, right=0.98, top=0.91, hspace=0.36)
fig.suptitle(dict_dataset_names[dataset])
ax = fig.add_subplot(2, 1, 1)
ax.plot(ts[extracted], avg_min_dists[extracted], '.-', label=r'avg. closest physical distance $d_{avg}$ ($m$)')
ax.plot(ts[extracted], min_min_dists[extracted], '.-', label=r'min. closest physical distance $d_{min}$ ($m$)')
# ax.plot(ts[extracted], violations[extracted], '.-', label='# of violating instances (count)')
ax.grid()
ax.set_xlabel(r'Time [$sec$]')
# ax.set_ylabel('Distance (m)')
ax.set(xlim=(0, t_max))
ax.legend(loc=1)
ax = fig.add_subplot(2, 1, 2)
ax.plot(ts[extracted], density[extracted], '.-', label=r'social density $\rho$ (ped./$m^2$)')
ax.grid()
ax.set_xlabel(r'Time [$sec$]')
# ax.set_ylabel('Density (ped./m^2)')
ax.set(xlim=(0, t_max))
ax.legend(loc=1)
fig.savefig(os.path.join(path_analysis, dataset + '_statistics_vs_time.png'))
plt.close(fig)
# ============= below are figures of 2d histograms ================
bin_size = 15
# ---- figure 4 ----
fig = plt.figure(figsize=(4.15, 3.27))
fig.subplots_adjust(left=0.18, right=1.0, top=0.9, bottom=0.14)
fig.suptitle(dict_dataset_names[dataset])
ax = fig.add_subplot(1, 1, 1)
plt.hist2d(avg_min_dists, density,
bins=(bin_size, bin_size),
# range=[[0.0, 6.0], [0.0, 0.15]],
# norm=LogNorm()
)
plt.colorbar()
ax.set_ylabel(r'Social Density $\rho$ (ped./$m^2$)')
ax.set_xlabel('Avg. Closest Physical Distance $d_{avg}$ ($m$)')
fig.savefig(os.path.join(path_analysis, dataset + '_2d_hist_density_vs_avg_dists.png'))
plt.close(fig)
# ---- figure 5 ----
fig = plt.figure(figsize=(4.15, 3.27))
fig.subplots_adjust(left=0.18, right=1.0, top=0.9, bottom=0.14)
fig.suptitle(dict_dataset_names[dataset])
ax = fig.add_subplot(1, 1, 1)
plt.hist2d(min_min_dists, density,
bins=(bin_size, bin_size),
# range=[[0.0, 6.0], [0.0, 0.15]],
# norm=LogNorm()
)
plt.colorbar()
ax.set_ylabel(r'Social Density $\rho$ (ped./$m^2$)')
ax.set_xlabel(r'Min. Closest Physical Distance $d_{min}$ ($m$)')
fig.savefig(os.path.join(path_analysis, dataset + '_2d_hist_density_vs_min_dists.png'))
plt.close(fig)
# ---- figure 6 ----
fig = plt.figure(figsize=(4.15, 3.27))
fig.subplots_adjust(left=0.18, right=1.0, top=0.9, bottom=0.14)
fig.suptitle(dict_dataset_names[dataset])
ax = fig.add_subplot(1, 1, 1)
plt.hist2d(violations, density,
bins=(bin_size, bin_size),
# range=[[0.0, 6.0], [0.0, 0.15]],
# norm=LogNorm()
)
plt.colorbar()
ax.set_ylabel(r'Social Density $\rho$ (ped./$m^2$)')
# ax.set_yticks(np.arange(0, 0.15, 0.01))
ax.set_xlabel(r'Num. of Social Distancing Violations $v$')
fig.savefig(os.path.join(path_analysis, dataset + '_2d_hist_density_vs_violation.png'))
plt.close(fig)
# ---- figure 7 ----
fig = plt.figure(figsize=(3.28, 3.14))
fig.subplots_adjust(left=0.21, bottom=0.15, right=0.97, top=0.90)
fig.suptitle(dict_dataset_names[dataset])
ax = fig.add_subplot(1, 1, 1)
ax.grid()
ax.plot(violations, density, '.b', alpha=0.2)
ax.set_ylabel(r'Social Density $\rho$ (ped./$m^2$)')
ax.set_xlabel(r'Num. of Social Distancing Violations $v$')
print('Avg. Avg. Closest Physical Distance = %.6f' % np.mean(avg_min_dists))
print('Skewness = %.6f' % stats.skew(density))
intercept, slope, preds, lbs, ubs, x_select, lb_select, ub_select = \
custom_simple_linear_regression(xs=violations, ys=density, x_select='y_intercept')
print('x_select = %.6f' % x_select)
print('x_select_lb = %.6f' % lb_select)
print('x_select_ub = %.6f' % ub_select)
line = [0, max(violations)]
# line = [-0.01, 0.175]
ax.plot(line, [intercept + slope * line[0], intercept + slope * line[1]], '-r')
ax.plot(preds, lbs, '-g')
ax.plot(preds, ubs, '-g')
# ax.plot(0.0, lb_select, '.r')
plt.text(0.0 + 0.5, lb_select - 0.005, r'$\rho_c$', fontsize=15, color='r')
ax.set(xlim=(0.0, np.max(violations)))
ax.set(ylim=(0.0, np.max(density)))
# ax.vlines(0, -0.01, 0.15)
# ax.hlines(0, line[0], line[1])
# w = 1/60*density_max
# ax.plot([-w, w], [ci[1], ci[1]], color='r')
# ax.plot([-w, w], [ci[0], ci[0]], color='r')
# ax.plot([0, 0], ci, color='r')
# ax.hlines(intercept - std_dev, -w, w, color='r')
# ax.vlines(0, intercept - std_dev, intercept + std_dev, color='r')
# plt.show()
fig.savefig(os.path.join(path_analysis, dataset + '_regression_density_vs_violation.png'))
plt.close(fig)
def main():
for dataset in ['oxford_town', 'mall', 'grand_central']:
# for dataset in ['oxford_town', 'mall']:
analyze_statistics(dataset=dataset)
if __name__ == '__main__':
main()