shivammalviya712 / trigger-word-detection Goto Github PK

can i use this model for multiple word detection ?
i want to train model for 5-6 trigger words . so i have to add all the wav files into positive folder right ? thats it ? will this work for multiple trigger words ?

• Couldn't find ffmpeg or avconv - defaulting to ffmpeg, but may not work
• Could not import the pyaudio c module '_portaudio'.
• ImportError: DLL load failed: The specified module could not be found.

Model is loaded
Realtime is done
start
yesssss
(array([ 0.00747681, 0.00985718, 0.01608276, ..., -0.00848389,
-0.01101685, -0.0112915 ]), 44100)
Traceback (most recent call last):
File "main.py", line 23, in
realtime.refresh_audio()
File "/home/dimanshu/Trigger-Word-Detection/code/realtime.py", line 41, in refresh_audio
self.new_x = self.spectrogram(self.new_audio).T
File "/home/dimanshu/Trigger-Word-Detection/code/realtime.py", line 63, in spectrogram
nchannels = sound.ndim
AttributeError: 'tuple' object has no attribute 'ndim'

error: this exceeds GitHub's file size limit of 100.00 MB

RuntimeWarning: divide by zero encountered in log10
Z = 10. * np.log10(spec)

how can i test my wav file that there is a trigger word in my file or not ? i don't want to use real time