forked from crowdbotp/OpenTraj
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathcrowd_density.py
229 lines (170 loc) · 8.18 KB
/
crowd_density.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
# Author: Pat Zhang
# Email: [email protected]
import sys
import os
import math
import numpy as np
import pandas as pd
import sympy as sp
import pickle
import seaborn as sns
import matplotlib.pyplot as plt
from scipy.spatial.distance import pdist,squareform
from toolkit.loaders.loader_eth import load_eth
from toolkit.core.trajdataset import TrajDataset
from toolkit.benchmarking.load_all_datasets import get_datasets, all_dataset_names,get_trajlets
from toolkit.core.trajlet import split_trajectories
from toolkit.benchmarking.utils.histogram_sampler import normalize_samples_with_histogram
from copy import deepcopy
def local_density(all_frames,trajlets,name):
#define local density function
#for all pedestrians at that time, find its distance to NN
distNN = []
dens_t = []
a=1
new_frames = []
for frame in all_frames:
if len(frame)>1:
#find pairwise min distance
distNN.append([])
dens_t.append([])
dist = squareform(pdist(frame[['pos_x','pos_y']].values))
pair_dist = []
for pi in dist:
pair_dist.append(np.array(pi))
min_pi = [j for j in pi if j>0.01]
if len(min_pi) == 0:
min_dist = 0.01
else:
min_dist = np.min(min_pi)
distNN[-1].append(min_dist)
#calculate local density for agent pj
for pj in range(len(dist)):
dens_t_i = 1/(2*np.pi)*np.sum(1/((a*np.array(distNN[-1]))**2)*np.exp(-np.divide((pair_dist[pj]**2),(2*(a*np.array(distNN[-1]))**2))))
dens_t[-1].append(dens_t_i)
frame.loc[frame.index[pj],'p_local'] = dens_t_i
new_frames.append(frame)
new_frames = pd.concat(new_frames)
new_traj = TrajDataset()
new_traj.data = new_frames
trajs = new_traj.get_trajectories(label="pedestrian")
trajlets[name] = split_trajectories(trajs, to_numpy=False)
#average local density for each trajlet
avg_traj_plocal=[]
for trajlet in trajlets[name]:
avg_traj_plocal.append(np.max(trajlet['p_local']))
return avg_traj_plocal
def global_density(all_frames,area):
#calculate global density as numebr of agents in the scene area at time t
frame_density_samples = []
new_frames = []
for frame in all_frames:
if len(frame)>0:
oneArea = area.loc[frame['scene_id'].values[0],'area']
frame_density_samples.append(len(frame) / oneArea)
return frame_density_samples
def run(datasets, output_dir):
all_names = ['ETH-Univ','ETH-Hotel','UCY-Zara','UCY-Univ','SDD-Coupa','SDD-bookstore','SDD-deathCircle','GC','InD-1','InD-2','KITTI','LCas-Minerva','WildTrack','Edinburgh','BN-1d-w180','BN-2d-w160']
#list(datasets.keys())
#store all the results in pandas dataframe
all_global_density=[]
all_local_density=[]
# Get trajectories from dataset
for ds_name in all_names:
dataset = datasets[ds_name]
all_frames = dataset.get_frames()
all_trajs = dataset.get_trajectories()
trajlets = {}
#get scene area
scenes_maxX = dataset.data.groupby(['scene_id'])['pos_x'].max()
scenes_minX = dataset.data.groupby(['scene_id'])['pos_x'].min()
scenes_maxY = dataset.data.groupby(['scene_id'])['pos_y'].max()
scenes_minY = dataset.data.groupby(['scene_id'])['pos_y'].min()
area=pd.DataFrame(data=[],columns=['scene_id','area'])
for idx in scenes_maxX.index:
x_range = scenes_maxX.loc[idx]-scenes_minX.loc[idx]
y_range = scenes_maxY.loc[idx]-scenes_minY.loc[idx]
area.loc[idx,'area'] = x_range*y_range
#calculate and store global density
global_dens = global_density(all_frames,area)
g_density = pd.DataFrame(data=np.zeros((len(global_dens),2)),columns=['ds_name','global_density'])
g_density.iloc[:,0] = [ds_name for i in range(len(global_dens))]
g_density.iloc[:,1] = global_dens
all_global_density.append(global_dens)
outputFile1 = output_dir+"/"+ds_name+'_globalDens.h5'
fw = open(outputFile1, 'wb')
pickle.dump(g_density, fw)
fw.close()
#calculate and store local density
trajlets = {}
local_dens = local_density(all_frames,trajlets,ds_name)
l_density = pd.DataFrame(data=[],columns=['ds_name','local_density'])
l_density.iloc[:,1] = local_dens
l_density.iloc[:,0] = [ds_name for i in range(len(l_density.iloc[:,1]))]
all_local_density.append(local_dens)
outputFile2 = output_dir+"/"+ds_name+'_localDens.h5'
fw = open(outputFile2, 'wb')
pickle.dump(l_density, fw)
fw.close()
print(ds_name," finish")
# down-sample each group.
# down-sample each group.
gdens_d = normalize_samples_with_histogram(all_global_density[:-2], max_n_samples=800, n_bins=50,quantile_interval=[0.05, 0.98])
ldens_d = normalize_samples_with_histogram(all_local_density[:-2],max_n_samples=800, n_bins=50,quantile_interval=[0.05, 0.95])
BNgdens_d = normalize_samples_with_histogram(all_global_density[-2:], max_n_samples=800, n_bins=50,quantile_interval=[0.05, 0.98])
BNldens_d = normalize_samples_with_histogram(all_local_density[-2:],max_n_samples=800, n_bins=50,quantile_interval=[0.05, 0.95])
# put samples in a DataFrame (required for seaborn plots)
df_gdens = pd.concat([pd.DataFrame({'title': all_names[ii],
'global_density': gdens_d[ii],
}) for ii in range(len(all_names[:-2]))])
df_ldens = pd.concat([pd.DataFrame({'title': all_names[ii],
'local_density': ldens_d[ii],
}) for ii in range(len(all_names[:-2]))])
BN_gdens = pd.concat([pd.DataFrame({'title': all_names[-ii-1],
'global_density': BNgdens_d[ii],
}) for ii in range(2)])
BN_ldens = pd.concat([pd.DataFrame({'title': all_names[-ii-1],
'local_density': BNldens_d[ii],
}) for ii in range(2)])
print("making plots ...")
sns.set(style="whitegrid")
fig,axs = plt.subplots(2,2,figsize=(12, 2),gridspec_kw={'width_ratios': [4, 1],'height_ratios': [1, 1]})
sns.swarmplot(y='global_density', x='title', data=df_gdens, size=1,ax=axs[0,0])
axs[0,0].set_ylim([0, 0.08])
axs[0,0].set_yticks([0, 0.02,0.04,0.06,0.08])
axs[0,0].set_xlabel('')
axs[0,0].yaxis.label.set_size(8)
axs[0,0].yaxis.set_tick_params(labelsize=8)
sns.swarmplot(y='local_density', x='title', data=df_ldens, size=1,ax=axs[1,0])
axs[1,0].set_ylim([0, 6])
axs[1,0].set_yticks([0, 2.0,4.0,6.0])
axs[1,0].yaxis.label.set_size(8)
axs[1,0].xaxis.set_tick_params(labelsize=8)
axs[1,0].set_xlabel('')
axs[1,0].tick_params(axis='x', labelrotation=-20)
axs[1,0].yaxis.set_tick_params(labelsize=8)
sns.swarmplot(y='global_density', x='title', data=BN_gdens, size=1,ax=axs[0,1])
axs[0,1].set_ylim([0, 1.5])
axs[0,1].set_yticks([0, 0.5,1,1.5])
axs[0,1].set_xlabel('')
axs[0,1].set_ylabel('')
axs[0,1].yaxis.set_tick_params(labelsize=8)
sns.swarmplot(y='local_density', x='title', data=BN_ldens, size=1,ax=axs[1,1])
axs[1,1].set_ylim([0, 6])
axs[1,1].set_yticks([0, 2,4,6])
axs[1,1].set_xlabel('')
axs[1,1].set_ylabel('')
axs[1,1].yaxis.set_tick_params(labelsize=8)
axs[1,1].xaxis.set_tick_params(labelsize=8)
plt.setp(axs[0,0].get_xticklabels(), visible=False)
plt.setp(axs[0,1].get_xticklabels(), visible=False)
fig.align_ylabels(axs[:, :])
plt.xticks(rotation=-20)
plt.subplots_adjust(hspace=0.18,wspace=0.12)
plt.savefig(os.path.join(output_dir, 'density.pdf'), dpi=500, bbox_inches='tight')
plt.show()
if __name__ == "__main__":
opentraj_root = sys.argv[1]
output_dir = sys.argv[2]
datasets = get_datasets(opentraj_root, all_dataset_names)
run(datasets, output_dir)