-
Notifications
You must be signed in to change notification settings - Fork 2
/
Copy pathDVS_Gesture_preprocess.py
169 lines (136 loc) · 6.05 KB
/
DVS_Gesture_preprocess.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
import os
import h5py
import struct
import glob
from DVS_dataload.events_timeslices import *
TimeStep = 40
ds = [4, 4]
dt = 30 * 1000
size=[2, 32, 32]
def gather_aedat(directory, start_id, end_id, filename_prefix='user'):
fns = []
for i in range(start_id, end_id):
search_mask = directory + os.sep + \
filename_prefix + "{0:02d}".format(i) + '*.aedat'
glob_out = glob.glob(search_mask)
if len(glob_out) > 0:
fns += glob_out
return fns
# read aedat
def aedat_to_events(filename):
# read label
label_filename = filename[:-6] + '_labels.csv'
labels = np.loadtxt(label_filename,
skiprows=1,
delimiter=',',
dtype='uint32')
events = []
with open(filename, 'rb') as f:
for i in range(5):
_ = f.readline()
while True:
data_ev_head = f.read(28)
if len(data_ev_head) == 0:
break
eventtype = struct.unpack('H', data_ev_head[0:2])[0]
eventsource = struct.unpack('H', data_ev_head[2:4])[0]
eventsize = struct.unpack('I', data_ev_head[4:8])[0]
eventoffset = struct.unpack('I', data_ev_head[8:12])[0]
eventtsoverflow = struct.unpack('I', data_ev_head[12:16])[0]
eventcapacity = struct.unpack('I', data_ev_head[16:20])[0]
eventnumber = struct.unpack('I', data_ev_head[20:24])[0]
eventvalid = struct.unpack('I', data_ev_head[24:28])[0]
if (eventtype == 1):
event_bytes = np.frombuffer(f.read(eventnumber * eventsize),
'uint32')
event_bytes = event_bytes.reshape(-1, 2)
x = (event_bytes[:, 0] >> 17) & 0x00001FFF
y = (event_bytes[:, 0] >> 2) & 0x00001FFF
p = (event_bytes[:, 0] >> 1) & 0x00000001
t = event_bytes[:, 1]
events.append([t, x, y, p])
else:
f.read(eventnumber * eventsize)
events = np.column_stack(events)
events = events.astype('uint32')
clipped_events = np.zeros([4, 0], 'uint32')
for l in labels:
start = np.searchsorted(events[0, :], l[1])
end = np.searchsorted(events[0, :], l[2])
clipped_events = np.column_stack([clipped_events,
events[:, start:end]])
return clipped_events.T, labels
# build hdf5
def sample(times, addrs, T=40, dt=30000, size=[2, 32, 32], ds=[4,4], is_train_Enhanced=False):
tbegin = times[0]
tend = np.maximum(0, times[-1] - T * dt)
start_time = np.random.randint(tbegin, tend) if is_train_Enhanced else 0
data = get_tmad_slice(times[()],
addrs[()],
start_time,
T * dt)
data[:, 0] -= data[0, 0]
t_start = data[0][0]
ts = range(t_start, t_start + T * dt, dt)
re = np.zeros([len(ts)] + size, dtype='int8')
idx_start = 0
idx_end = 0
for i, t in enumerate(ts):
idx_end += find_first(data[idx_end:, 0], t + dt)
if idx_end > idx_start:
data_temp = data[idx_start:idx_end, 1:]
pol, y, x = data_temp[:, 2], (data_temp[:, 0] // ds[0]).astype(np.int),\
(data_temp[:, 1] // ds[1]).astype(np.int)
np.add.at(re, (i, pol, x, y), 1)
idx_start = idx_end
return re
def create_hdf5(path, save_path):
print('processing train data...')
save_path_train = os.path.join(save_path, 'DvsGesture_train_40step_downsample')
if not os.path.exists(save_path_train):
os.makedirs(save_path_train)
fns_train = gather_aedat(path, 1, 24)
index_data_save = 0
for i in range(len(fns_train)):
print('processing training data: {}/{}, {:.1f} %'.format(i+1, len(fns_train), 100.*(i+1)/len(fns_train)))
data, labels_starttime = aedat_to_events(fns_train[i])
tms = data[:, 0]
ads = data[:, 1:]
lbls = labels_starttime[:, 0]
start_tms = labels_starttime[:, 1]
end_tms = labels_starttime[:, 2]
for lbls_idx in range(len(lbls)):
s_ = get_slice(tms, ads, start_tms[lbls_idx], end_tms[lbls_idx])
data = sample(s_[0], s_[1], T=TimeStep, dt=dt, size=size, ds=ds, is_train_Enhanced=False)
index_data_save += 1
with h5py.File(save_path_train + os.sep + 'DVS-Gesture-train' + str(index_data_save) + '.hdf5',
'w') as f:
f.create_dataset('data', data=data, dtype=np.int8)
f.create_dataset('label', data=lbls[lbls_idx] - 1, dtype=np.int8)
print('Training data processing completed')
print('processing test data...')
save_path_test = os.path.join(save_path, 'DvsGesture_test_40step_downsample')
if not os.path.exists(save_path_test):
os.makedirs(save_path_test)
fns_test = gather_aedat(path, 24, 30)
index_data_save = 0
for i in range(len(fns_test)):
print('processing testing data: {}/{}, {:.1f} %'.format(i+1, len(fns_test), 100.*(i+1)/len(fns_test)))
data, labels_starttime = aedat_to_events(fns_test[i])
tms = data[:, 0]
ads = data[:, 1:]
lbls = labels_starttime[:, 0]
start_tms = labels_starttime[:, 1]
end_tms = labels_starttime[:, 2]
for lbls_idx in range(len(lbls)):
s_ = get_slice(tms, ads, start_tms[lbls_idx], end_tms[lbls_idx])
data = sample(s_[0], s_[1], T=TimeStep, dt=dt, size=size, ds=ds, is_train_Enhanced=False)
index_data_save += 1
with h5py.File(save_path_test + os.sep + 'DVS-Gesture-test' + str(index_data_save) + '.hdf5',
'w') as f:
f.create_dataset('data', data=data, dtype=np.uint8)
f.create_dataset('label', data=lbls[lbls_idx] - 1, dtype=np.uint8)
print('Test data processing completed')
if __name__ == '__main__':
path = os.getcwd() + os.sep + 'data' + os.sep + 'DVS_Gesture'
create_hdf5(os.path.join(path, 'source_DvsGesture/DvsGesture'), path)