-
Notifications
You must be signed in to change notification settings - Fork 3
/
simulate_anechoic.py
187 lines (135 loc) · 7.22 KB
/
simulate_anechoic.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
import math
import random
import numpy as np
import prepare_data as pp
import os
from dab import dab_run
import metrics as m
import sys
def query_yes_no(question, default="yes"):
"""Ask a yes/no question via raw_input() and return their answer.
"question" is a string that is presented to the user.
"default" is the presumed answer if the user just hits <Enter>.
It must be "yes" (the default), "no" or None (meaning
an answer is required of the user).
The "answer" return value is True for "yes" or False for "no".
"""
valid = {"yes": True, "y": True, "ye": True,
"no": False, "n": False}
if default is None:
prompt = " [y/n] "
elif default == "yes":
prompt = " [Y/n] "
elif default == "no":
prompt = " [y/N] "
else:
raise ValueError("invalid default answer: '%s'" % default)
while True:
sys.stdout.write(question + prompt)
choice = input().lower()
if default is not None and choice == '':
return valid[default]
elif choice in valid:
return valid[choice]
else:
sys.stdout.write("Please respond with 'yes' or 'no' "
"(or 'y' or 'n').\n")
def set_microphone_at_distance(clean_data, noise_data, framerate, distance):
meter_snr = 10 ** (15 / 20) # chosen snr at one meter distance (15 dB)
clean_energy = 0
noise_energy = 0
sound_speed = 343
frame_delay = int(math.ceil(framerate * distance / sound_speed)) # compute number of frame to delay
delay_silence = np.zeros(frame_delay)
clean_data = np.append(delay_silence, clean_data) # add delay to attenuated speech
shift = random.randint(0, abs(len(clean_data) - len(
noise_data))) # calculate random shift for noise in the possible range, to avoid "overflows"
n = min(len(clean_data), len(noise_data))
clean_data = clean_data[0:n]
noise_data = noise_data[shift:(n + shift)]
for t in clean_data:
clean_energy = clean_energy + abs(t)
for t in noise_data:
noise_energy = noise_energy + abs(t)
first_snr = clean_energy / noise_energy
snr_ratio = meter_snr / first_snr
new_noise_data = noise_data / snr_ratio # normalizing snr level at 15 db at one meter distance
new_noise_energy = noise_energy/snr_ratio # updating noisy energy after normalization
# mixed_data = VectorAdd(noise_data, clean_data/distance) # attenuating clean speech at 1/distance rate and adding noise
mixed_data = (new_noise_data + (clean_data / distance)) # attenuating clean speech at 1/distance rate and adding noise
final_s2nr = (clean_energy/distance)/(new_noise_energy + clean_energy/distance) #calculating s2nr of attenuated speech
return mixed_data, new_noise_data, clean_data, final_s2nr
def create_room(source_file, noise_file, dist):
(clean, fs) = pp.read_audio(source_file)
(noise, _) = pp.read_audio(noise_file)
for file in os.listdir(os.path.join("data_eval", "dnn1_in")):
file_path = os.path.join("data_eval", "dnn1_in", file)
os.remove(file_path)
for n in range(len(dist)):
mixed, noise_new, clean_new, s2nr = set_microphone_at_distance(clean, noise, fs, dist[n])
# s2nr = 1 / (1 + (1 / float(snr)))
mixed_name = "mix_%s_%s" % (str(dist[n]), os.path.basename(source_file))
clean_name = "clean_%s_%s" % (str(dist[n]), os.path.basename(source_file))
mixed_path = os.path.join('data_eval/dnn1_in', mixed_name)
clean_path = os.path.join('data_eval/dnn1_in', clean_name)
pp.write_audio(mixed_path, mixed, fs)
#pp.write_audio(clean_path, clean_new, fs)
go_on = True
while go_on == True:
work_dir = 'data_eval'
speech_dir = os.path.join('data_eval', 'test_speech')
input_string = input('Insert microphones distances: ')
dist = input_string.split()
dist = list(map(int, dist))
d_mean = math.ceil(sum(dist)/len(dist))
snr = [5.62/d for d in dist]
source_files = [f for f in os.listdir(speech_dir)
if f.endswith(".wav")]
# executing for all source speeches in speech folder
for f in source_files:
create_room(os.path.join(speech_dir, f), 'noise/babble.wav', dist)
dab_run(snr, f, mode='dab')
dab_run(snr, f, mode='db')
print("--------NOiSY----------------------------------\n -------------")
# calculating speech-per-speech indexes
m.calculate_pesq_couple(speech_dir, 'data_eval/dnn1_in')
avg_pesqs_N, std_pesqs_N = m.get_pesq_stats()
avg_stoi_N, std_stoi_N = m.calc_stoi_couple(speech_dir, 'data_eval/dnn1_in')
avg_sdr_N, std_sdr_N = m.calc_sdr_couple(speech_dir, 'data_eval/dnn1_in')
print("--------DS-------------------------------------\n -------------")
# calculating speech-per-speech indexes
m.calculate_pesq_couple(speech_dir, 'data_eval/dnn1_out')
avg_pesqs_DS, std_pesqs_DS = m.get_pesq_stats()
avg_stoi_DS, std_stoi_DS = m.calc_stoi_couple(speech_dir, 'data_eval/dnn1_out')
avg_sdr_DS, std_sdr_DS = m.calc_sdr_couple(speech_dir, 'data_eval/dnn1_out')
print("--------DB-------------------------------------\n -------------")
m.calculate_pesq_couple(speech_dir, 'data_eval/db')
avg_pesqs_DB, std_pesqs_DB = m.get_pesq_stats()
avg_stoi_DB, std_stoi_DB = m.calc_stoi_couple(speech_dir, 'data_eval/db')
avg_sdr_DB, std_sdr_DB = m.calc_sdr_couple(speech_dir, 'data_eval/db')
print("--------DAB------------------------------------\n -------------")
# calculating speech-per-speech indexes
m.calculate_pesq_couple(speech_dir, 'data_eval/dab')
avg_pesqs_DAB, std_pesqs_DAB = m.get_pesq_stats()
avg_stoi_DAB, std_stoi_DAB = m.calc_stoi_couple(speech_dir, 'data_eval/dab')
avg_sdr_DAB, std_sdr_DAB = m.calc_sdr_couple(speech_dir, 'data_eval/dab')
print('-----------------------------------------------------------------------------------------------------------')
print('INDEX:\t STOI\t PESQ\t SDR\t ---------------------------------------------------------------------------')
print('Noisy \t %f(%f) \t %f(%f) \t %f(%f)' % (avg_stoi_N, std_stoi_N,
avg_pesqs_N, std_pesqs_N,
avg_sdr_N, std_sdr_N))
print('DS \t %f(%f) \t %f(%f) \t %f(%f)' % (avg_stoi_DS, std_stoi_DS,
avg_pesqs_DS, std_pesqs_DS,
avg_sdr_DS, std_sdr_DS))
print('DB \t %f(%f) \t %f(%f) \t %f(%f)' % (avg_stoi_DB, std_stoi_DB,
avg_pesqs_DB, std_pesqs_DB,
avg_sdr_DB, std_sdr_DB))
print('DAB \t %f(%f) \t %f(%f) \t %f(%f)' % (avg_stoi_DAB, std_stoi_DAB,
avg_pesqs_DAB, std_pesqs_DAB,
avg_sdr_DAB, std_sdr_DAB))
room_dims = [30, 30]
iterations = 10000
steps = 1000
n_mics = len(dist)
# distrib = m.monte_carlo(room_dims, n_mics, iterations, steps, 'rect')
go_on = query_yes_no('Simulate new room?')