-
Notifications
You must be signed in to change notification settings - Fork 40
/
dataset.py
executable file
·179 lines (160 loc) · 6.1 KB
/
dataset.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
from torch.utils.data.dataset import Dataset
import torch
import os
import pickle
from liegroups.torch import SO3
class KAISTDataset(Dataset):
def __init__(self, args):
self.name = args.dataset_name
self.path_data_save = args.path_data_save
self.path_results = args.path_results
self.path_temp = args.path_temp
self.test_sequences = args.test_sequences
self.cross_validation_sequences = args.cross_validation_sequences
self.get_datasets()
self.set_normalize_factors(args)
# Transformation frame for Kaist dataset
T_vehicle2fog = torch.eye(4)
T_vehicle2imu = torch.eye(4)
T_vehicle2fog[:3, 3] = torch.Tensor([-0.335, -0.035, 0.78])
T_vehicle2imu[:3, 3] = torch.Tensor([-0.07, 0, 1.7])
self.calibration_parameters = {"Encoder resolution": 4096,
"Encoder left wheel diameter": 0.623803,
"Encoder right wheel diameter": 0.623095,
"Encoder wheel base": 1.52683,
"Vehicle2FOG": T_vehicle2fog,
"Vehicle2IMU": T_vehicle2imu}
def get_datasets(self):
self.datasets = []
for dataset in os.listdir(self.path_data_save):
self.datasets += [dataset[:-2]] # take just name
self.divide_datasets()
def divide_datasets(self):
self.datasets_test = self.test_sequences
self.datasets_validation = self.cross_validation_sequences
self.datasets_train = []
for dataset in self.datasets:
if (not dataset in self.datasets_test) and (not dataset in self.datasets_validation):
self.datasets_train += [dataset]
def dataset_name(self, i):
return self.datasets[i]
def get_filter_data(self, i):
if type(i) != int:
i = self.datasets.index(i)
pickle_dict = self[i]
t = pickle_dict['t']
chi0 = pickle_dict['chi'][0]
Rot0 = chi0[:3, :3]
angles = SO3.from_matrix(Rot0).to_rpy()
p0 = chi0[:3, 3]
u_odo_fog = pickle_dict['u_odo_fog']
y_imu = pickle_dict['u_imu']
x0 = torch.zeros(9)
x0[:3] = p0
x0[3:6] = angles
return t, x0, u_odo_fog, y_imu
def get_ground_truth_data(self, i):
pickle_dict = self[self.datasets.index(i) if type(i) != int else i]
return pickle_dict['t'], pickle_dict['chi']
def get_test_data(self, i, gp_name):
var = "odo_fog" if gp_name == "GpOdoFog" else "imu"
dataset = self.datasets_test[i] if type(i) == int else i
pickle_dict = self[self.datasets.index(dataset)]
u = pickle_dict["u_" + var]
y = pickle_dict["y_" + var]
u = self.normalize(u, "u_" + var)
if u[0].norm() == 0: #(Urban00-05 and campus00)
u = torch.zeros(0, u.shape[1], u.shape[2])
y = torch.zeros(0, y.shape[1])
return u, y
def get_validation_data(self, i, gp_name):
var = "odo_fog" if gp_name == "GpOdoFog" else "imu"
dataset = self.datasets_validation[i] if type(i) == int else i
pickle_dict = self[self.datasets.index(dataset)]
u = pickle_dict["u_" + var]
y = pickle_dict["y_" + var]
u = self.normalize(u, "u_" + var)
if u[0].norm() == 0: # (Urban00-05 and campus00)
u = torch.zeros(0, u.shape[1], u.shape[2])
y = torch.zeros(0, y.shape[1])
return u, y
def get_train_data(self, i, gp_name):
var = "odo_fog" if gp_name == "GpOdoFog" else "imu"
dataset = self.datasets_train[i] if type(i) == int else i
pickle_dict = self[self.datasets.index(dataset)]
u = pickle_dict["u_" + var]
y = pickle_dict["y_" + var]
u = self.normalize(u, "u_" + var)
if u[0].norm() == 0: # (Urban00-05 and campus00)
u = torch.zeros(0, u.shape[1], u.shape[2])
y = torch.zeros(0, y.shape[1])
return u, y
def __getitem__(self, i):
with open(self.path_data_save + self.datasets[i] + '.p', "rb") as file_pi:
mondict = pickle.load(file_pi)
return mondict
def __len__(self):
return len(self.datasets)
def set_normalize_factors(self, args):
"""
Compute mean and variance of input data using only training data
"""
# first mean
self.num_data = 0
for i, dataset in enumerate(self.datasets_train):
with open(self.path_data_save + dataset + '.p', "rb") as file_pi:
pickle_dict = pickle.load(file_pi)
u_odo_fog = pickle_dict['u_odo_fog']
u_imu = pickle_dict['u_imu']
if i == 0:
u_odo_fog_loc = u_odo_fog.mean(dim=0).mean(dim=0)
u_imu_loc = u_imu.mean(dim=0).mean(dim=0)
else:
u_odo_fog_loc += u_odo_fog.mean(dim=0).mean(dim=0)
u_imu_loc += u_imu.mean(dim=0).mean(dim=0)
self.num_data += u_imu.shape[0]
u_odo_fog_loc = u_odo_fog_loc/len(self.datasets_train)
u_imu_loc = u_imu_loc/len(self.datasets_train)
# second standard deviation
u_length = 0
for i, dataset in enumerate(self.datasets_train):
with open(self.path_data_save + dataset + '.p', "rb") as file_pi:
pickle_dict = pickle.load(file_pi)
u_odo_fog = pickle_dict['u_odo_fog']
u_imu = pickle_dict['u_imu']
if i == 0:
u_odo_fog_std = ((u_odo_fog-u_odo_fog_loc)**2).sum(dim=0).sum(dim=0)
u_imu_std = ((u_imu-u_imu_loc)**2).sum(dim=0).sum(dim=0)
else:
u_odo_fog_std += ((u_odo_fog - u_odo_fog_loc)**2).sum(dim=0).sum(dim=0)
u_imu_std += ((u_imu - u_imu_loc)**2).sum(dim=0).sum(dim=0)
u_length += u_odo_fog.shape[0]*u_odo_fog.shape[1]
u_odo_fog_std = (u_odo_fog_std/u_length).sqrt()
u_imu_std = (u_imu_std/u_length).sqrt()
# for constant measurements, set standard deviation to 1
u_odo_fog_std[u_odo_fog_std == 0] = 1
u_imu_std[u_imu_std == 0] = 1
self.normalize_factors = {
'u_odo_fog_loc': u_odo_fog_loc,
'u_imu_loc': u_imu_loc,
'u_odo_fog_std': u_odo_fog_std,
'u_imu_std': u_imu_std,
}
pickle_dict = {'normalize_factors': self.normalize_factors}
with open(self.path_temp + "normalize_factors.p", "wb") as file_pi:
pickle.dump(pickle_dict, file_pi)
def normalize(self, x, var="u_odo_fog"):
x_loc = self.normalize_factors[var + "_loc"]
x_std = self.normalize_factors[var + "_std"]
x_normalized = (x-x_loc)/x_std
return x_normalized
class NCLTDataset(KAISTDataset):
def __init__(self, args):
super(NCLTDataset, self).__init__(args)
# Transformation frame for NCLT dataset
T_vehicle2fog = torch.eye(4)
T_vehicle2imu = torch.eye(4)
T_vehicle2fog[:3, 3] = torch.Tensor([0, -0.25, -0.49])
T_vehicle2imu[:3, 3] = torch.Tensor([-0.11, -0.18, -0.71])
self.calibration_parameters = {"Vehicle2FOG": T_vehicle2fog,
"Vehicle2IMU": T_vehicle2imu}