-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathmain_GUI.py
192 lines (166 loc) · 7.13 KB
/
main_GUI.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
#!/usr/bin/python
# -*- coding: utf8 -*-
from Tkinter import *
from features import mfcc
from features import logfbank
from DTW import *
import scipy.io.wavfile as wav
import pyaudio
import wave
import cPickle
import glob
import os
import argparse
CHUNK = 1024
FORMAT = pyaudio.paInt16
CHANNELS = 2
RATE = 44100
class Application(Frame):
def __init__(self, master=None):
Frame.__init__(self, master)
self.grid()
self.createWidgets()
self.record_seconds = 3 #'''每筆音訊都是三秒'''
self.no_cluster = False
self.hac = False
self.k_means = False
self.label_feat = [] #'''事先錄好的Feature Signal'''
self.prediction = "安安你好"
self.dtw = DTW()
def record(self):
#start Recording...
self.record_seconds = int(self.second_entry.get())
if self.file_entry.get() == "":
filename = "output"
else:
filename = self.file_entry.get()
p = pyaudio.PyAudio()
stream = p.open(format=FORMAT,
channels=CHANNELS,
rate=RATE,
input=True,
frames_per_buffer=CHUNK)
frames = []
for i in range(0, int(RATE / CHUNK * self.record_seconds)):
#if i % 60 == 0:
# self.counter_label.config(text="%i" % i)
print i / 42.0
data = stream.read(CHUNK)
frames.append(data)
stream.stop_stream()
stream.close()
p.terminate()
wf = wave.open("model_%s/%s.wav" % (SETS,filename), 'wb')
wf.setnchannels(CHANNELS)
wf.setsampwidth(p.get_sample_size(FORMAT))
wf.setframerate(RATE)
wf.writeframes(b''.join(frames))
wf.close()
#MFCC_dimension = 13
(rate,sig) = wav.read("model_%s/%s.wav" % (SETS,filename))
#Dump MFCC pickle file
mfcc_feat = mfcc(sig,rate)
cPickle.dump(mfcc_feat, open("model_%s/%s.mfcc" % (SETS,filename), "wb"))
#Dump HAC pickle file
hac_feat = self.dtw.calc_HAC(mfcc_feat)
cPickle.dump(hac_feat, open("model_%s/%s.hac" % (SETS,filename), "wb"))
#Dump K-means pickle file
k_means_feat = self.dtw.calc_Kmeans(mfcc_feat)
cPickle.dump(k_means_feat, open("model_%s/%s.kmeans" % (SETS,filename), "wb"))
#Pridict or Just Record only
if self.no_cluster or self.hac or self.k_means:
start_time = time.time()
#TODOs: Load MFCC Label Feature (pkl files) as self.label_feat
self.label_feat = []
if not self.hac and not self.k_means:
list_of_files = glob.glob('model_%s/*.mfcc' % SETS)[1:]
for filename in list_of_files:
self.label_feat.append(cPickle.load( open( filename, "rb" ) ))
dtw_result = [ self.dtw.calc_DTW(mfcc_feat, arr2) for arr2 in self.label_feat]
# for i in range(len(list_of_files)):
# print list_of_files[i], ":", dtw_result[i]
elif self.hac and not self.k_means:
print "HAC"
list_of_files = glob.glob('model_%s/*.hac' % SETS)[1:]
for filename in list_of_files:
self.label_feat.append(cPickle.load( open( filename, "rb" ) ))
dtw_result = [ self.dtw.calc_DTW(hac_feat, arr2) for arr2 in self.label_feat]
# for i in range(len(list_of_files)):
# print list_of_files[i], ":", dtw_result[i]
elif self.k_means and not self.hac:
print "K-means"
list_of_files = glob.glob('model_%s/*.kmeans' % SETS)[1:]
for filename in list_of_files:
self.label_feat.append(cPickle.load( open( filename, "rb" ) ))
dtw_result = [ self.dtw.calc_DTW(k_means_feat[0], arr2[0]) for arr2 in self.label_feat]
# for i in range(len(list_of_files)):
# print list_of_files[i], ":", dtw_result[i]
self.prediction = str(list_of_files[dtw_result.index(min(dtw_result))].split(".")[0][9:])
self.result_label.config(text='%s' % self.prediction, fg="red")
elapsed_time = time.time() - start_time
self.counter_label.config(text="DTW時間: %.2f" % elapsed_time)
os.remove("model_%s/%s.wav" % (SETS, self.file_entry.get()))
os.remove("model_%s/%s.mfcc" % (SETS, self.file_entry.get()))
os.remove("model_%s/%s.hac" % (SETS, self.file_entry.get()))
os.remove("model_%s/%s.kmeans" % (SETS, self.file_entry.get()))
def modeSelect(self, mode):
if self.no_cluster == False and mode == "mfcc" :
self.no_cluster = True
self.hac = False
self.k_means = False
else:
self.no_cluster = False
if self.hac == False and mode == "hac":
print "HAC"
self.no_cluster = False
self.hac = True
self.k_means = False
else:
self.hac = False
if self.k_means == False and mode == "k-means":
print "k_means"
self.no_cluster = False
self.hac = False
self.k_means = True
else:
self.k_means = False
def createWidgets(self):
self.file_label = Label(self, text="檔名：")
self.file_label.grid(row=0, column=0)
self.file_entry = Entry(self)
self.file_entry.grid(row=0, column=1)
self.second_label = Label(self, text="秒數：")
self.second_label.grid(row=1, column=0)
self.second_entry = Entry(self)
self.second_entry.grid(row=1, column=1)
self.mode_check = Checkbutton(self, text="DTW", command=lambda: self.modeSelect("mfcc"))
self.mode_check.grid(row=0, column = 2)
self.mode_check = Checkbutton(self, text="DTW(HAC)", command=lambda: self.modeSelect("hac"))
self.mode_check.grid(row=1, column = 2)
self.mode_check = Checkbutton(self, text="DTW(K-means)", command=lambda: self.modeSelect("k-means"))
self.mode_check.grid(row=2, column = 2)
self.record_label = Label(self, text="辨識結果：")
self.record_label.grid(row=2, column=0)
self.result_label = Label(self, text="安安你好")
self.result_label.grid(row=2, column=1)
self.counter_label = Label(self, text="DTW時間: %.2f" % 0.0)
self.counter_label.grid(row=4, column=1)
self.Record = Button(self, text="錄音")
self.Record["fg"] = "pink"
self.Record["command"] = self.record
self.Record.grid(row = 3, column = 0)
self.Quit = Button(self, text="離開")
self.Quit["fg"] = "red"
self.Quit["bg"] = "blue"
self.Quit["command"] = self.quit
self.Quit.grid(row = 3, column = 1)
parser = argparse.ArgumentParser()
parser.add_argument('-cluster', type=str, default='15')
args = parser.parse_args()
SETS = args.cluster
root = Tk()
app = Application(master=root)
app.master.title('Dynamic Time Warping')
app.master.maxsize(1000,400)
app.mainloop()
root.destroy()