C:\Users\PC\PycharmProjects\untitled\Net resource\vad.py:150: RuntimeWarning: invalid value encountered in double_scalars
speech_ratio = sum_voice_energy / sum_full_energy

The above is an error that occurred when I ran the following code. How can I solve this problem？

from vad import VoiceActivityDetector
filename = r'C:\Users\PC\PycharmProjects\untitled\audio\test.wav'
v = VoiceActivityDetector(filename)
v.plot_detected_speech_regions()

Hello! For-loop in this function bugs my brain.

def _connect_energy_with_frequencies(self, data_freq, data_energy):
    energy_freq = {}
    for (i, freq) in enumerate(data_freq):
        energy_freq[abs(freq)] = data_energy[i]
    return energy_freq

data_freq array is "symmetrical" such that if there is '-1.3' there also is '1.3'.
We assigning different values to same dict key twice [abs(freq)] .
Not only this is pointless from general performance perspective.
Sign-opposite keys in dict represent amplitudes of same frequency wave in antiphase.
Shouldn't we do this instead ?

def _connect_energy_with_frequencies(self, data_freq, data_energy):
    energy_freq = {}
    for (i, freq) in enumerate(data_freq):
        if abs(freq) in energy_freq:
            energy_freq[abs(freq)] += data_energy[i]
        else:
            energy_freq[abs(freq)] = data_energy[i]
    return energy_freq

Or perhaps '-=' ?

你想用 ratio ，却写成了 ration in：

Basic Idea
Input audio data treated as following:

• Convert stereo to mono
• Move a window of 20ms along the audio data
• Calculate ration between energy of speech band and total energy for window
• If ratio is more than threshold (0.6 by default) label windows as speech
• Apply median filter with length of 0.5s to smooth detected speech regions
• Represent speech regions as intervals of time

Hi. Is there a reason why VAD would fail to identify speech in foreign language in some cases? I tested it using some audio in spanish and french, and the convert_windows_to_readible_labels() failed to return any labels for those audio. I was wondering if there were some parameters I needed to initialize in those cases

I am trying to adapt this script to detect voice-silence segments in an audio file containing source separated singing voice signal obtained from http://github.com/sigsep/open-unmix-pytorch

I have some questions:

• Does it make sense to compute the threshold for each data_window independently? Instead of having a fixed speech_energy_threshold? I would do that by computing the energy of the data_window signal, normalizing it and taking its mean value. If this value is = 0.0, I can label that segment as silence.

• Is there a clever way to choose parameters like sample_window, sample_overlap, speech_window that would be more appropriate for singing voice signals?

Thanks a lot!

How can I get data incides where speech starts or ends, instead of time when the same happens?

I recently just searched the term "VAD" in github search and found many abandoned projects with some decent amount of traction, but mostly lacking pre-trained models.

So I decided to share our new pre-trained VAD:

• Please see this vad - https://github.com/snakers4/silero-vad
• Quality benchmarks here - https://github.com/snakers4/silero-vad#vad-quality-metrics
• Overall comparison with webrtc - wiseman/py-webrtcvad#68

Hello,
I am radio amateur and trying to find solution to detect voice in VHF audio record.
I have this sample without voice, but you algorithm detect it.

the result.json is none

I had an issue when i am running the code; it says:
Traceback (most recent call last):
File "detectVoiceInWave.py", line 6, in
v = VoiceActivityDetector(filename)
File "/home/ebtesam/Desktop/Code/VAD-python-master/VAD-python-master/vad.py", line 11, in init
self._read_wav(wave_input_filename)._convert_to_mono()
File "/home/ebtesam/Desktop/Code/VAD-python-master/VAD-python-master/vad.py", line 20, in _read_wav
self.rate, self.data = wf.read(wave_file)
File "/home/ebtesam/.local/lib/python3.7/site-packages/scipy/io/wavfile.py", line 246, in read
raise ValueError("Unexpected end of file.")
ValueError: Unexpected end of file.

how can i solve it?

When attempted, throws the error:

"ValueError: File format b'fLaC'... not understood."

Hello,

is there also an option to run this VAD on a frame (e.g. 30ms) of an microphone input signal?
Would be nice to provide an example script using PyAudio and this module.

Regards,
Josef