-
Notifications
You must be signed in to change notification settings - Fork 1
/
Testing_iterative_alphabetical_clean_recordings.py
188 lines (126 loc) · 5.03 KB
/
Testing_iterative_alphabetical_clean_recordings.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
import os
import _pickle as pickle
import numpy as np
import scipy.io.wavfile as wav
from python_speech_features import mfcc
import Record_audio_Fixed_Interval as recor
import pandas as pd
import librosa
import logmmse
#import warnings
#warnings.filterwarnings("ignore")
import time
import sys
sys.path.append(os.path.abspath("C:/Users/panch/Desktop/Version3/libs/filters/"))
sys.path.append(os.path.abspath("C:/Users/panch/Desktop/Version3/libs/feature/"))
import silence as silence_reduction
#path to training data
source = "Dataset/"
#path where training speakers will be saved
modelpath = "Models/"
for charact_no in range(97,111):
#user = input("Enter user you want to test: ")
#i = 1
user = chr(charact_no)
i=1
ii=5
#ii=int(input("Enter No of Tests: "))
fs, sig = wav.read(source+user+"/"+"Testing/"+user+" ("+str(i)+").wav" )
#audio= silence_reduction.remove_silence(fs,sig)
#audio = logmmse.logmmse(audio,fs)
#sig=audio
file_name = user+" ("+str(i)+")"
vector = mfcc(sig,fs)
vector_df = pd.DataFrame(vector)
#vector = librosa.feature.mfcc(sig,fs)
mfcc_delta = librosa.feature.delta(vector)
mfcc_delta_df = pd.DataFrame(mfcc_delta)
mfcc_delta2 = librosa.feature.delta(vector, order=2)
mfcc_delta2_df = pd.DataFrame(mfcc_delta2)
Features = pd.concat([vector_df, mfcc_delta_df,mfcc_delta2_df], axis=1)
FINAL_FEATURES = pd.DataFrame(Features)
FINAL_FEATURES.set_axis(list(range(0, 39)), axis='columns', inplace=True)
Feature_DF = pd.DataFrame(FINAL_FEATURES)
#vector = librosa.feature.mfcc(y=sig,sr=fs,n_mfcc=40)
#vector_df = pd.DataFrame(vector)
#vector = mfcc(audio,fs)
#vector_df = pd.DataFrame(vector)
#mfcc_delta = librosa.feature.delta(vector)
#mfcc_delta_df = pd.DataFrame(mfcc_delta)
#mfcc_delta2 = librosa.feature.delta(vector, order=2)
#mfcc_delta2_df = pd.DataFrame(mfcc_delta2)
#Features = pd.concat([vector_df, mfcc_delta_df,mfcc_delta2_df], axis=1)
#vector_df = pd.DataFrame(Features)
#vector_df.set_axis(list(range(0, 39)), axis='columns', inplace=True)
#vector_df.to_csv("Vector_Test.csv")
File_array = []
Score_array = []
Max_Array =["Max_Score"]
path = "Models/"
dirs = os.listdir(path)
Testing_dir = source+user+"/"+"Testing/"
for fileas in dirs:
File_array.append(fileas)
trained_model = open(path+fileas,"rb")
gmm_Model = pickle.load(trained_model)
score = gmm_Model.score(Feature_DF)
#prob = np.exp(score)
Score_array = np.append(Score_array,score)
#Prob_Array = np.append(Prob_Array,prob)
File_array = pd.Series(File_array)
Score_array = pd.Series(Score_array)
Max_value = max(Score_array)
Max_Array.append(Max_value)
#Min_array = pd.Series(Prob_Array)
Df = pd.DataFrame([File_array,Score_array])
#--------------------------------------------------
i=i+1
while i<ii:
fs, sig = wav.read(source+user+"/"+"Testing/"+user+" ("+str(i)+").wav" )
audio= silence_reduction.remove_silence(fs,sig)
audio = logmmse.logmmse(audio,fs)
#audio=sig
#silence_reduction.remove_silence(sr,audio)
#audio = logmmse.logmmse_from_file(source+user+"/"+"Training/"+user+" ("+str(i)+").wav")
sig=audio
file_name = user+" ("+str(i)+")"
vector = mfcc(sig,fs)
vector_df = pd.DataFrame(vector)
#vector = librosa.feature.mfcc(sig,fs)
mfcc_delta = librosa.feature.delta(vector)
mfcc_delta_df = pd.DataFrame(mfcc_delta)
mfcc_delta2 = librosa.feature.delta(vector, order=2)
mfcc_delta2_df = pd.DataFrame(mfcc_delta2)
Features = pd.concat([vector_df, mfcc_delta_df,mfcc_delta2_df], axis=1)
FINAL_FEATURES = pd.DataFrame(Features)
FINAL_FEATURES.set_axis(list(range(0, 39)), axis='columns', inplace=True)
Feature_DF = pd.DataFrame(FINAL_FEATURES)
Score_array = []
for fileas in dirs:
#File_array.append(fileas)
trained_model = open(path+fileas,"rb")
gmm_Model = pickle.load(trained_model)
score = gmm_Model.score(FINAL_FEATURES)
#prob = np.exp(score)
Score_array = np.append(Score_array,score)
#Prob_Array = np.append(Prob_Array,prob)
Score_array = pd.Series(Score_array)
Max_value = max(Score_array)
Max_Array.append(Max_value)
Df=Df.append(Score_array,ignore_index=True)
i=i+1
mdf=pd.DataFrame(Max_Array)
Df = pd.concat([Df, mdf], axis=1)
Df.columns = Df.iloc[0]
Df = Df[1:]
#print(Df)
Max = Df["Max_Score"]
Current = Df[str(user)+".gmm"]
#print(Max)
#print("11111111111111111111")
#print(Current)
new_array = np.intersect1d(Max,Current)
Correct=len(new_array)
accuracy = (Correct/(ii-1))*100
print(user,accuracy)
Df.to_csv("Results/"+user+".csv")