webrtcvad_voice_activity_detection_on_silent_audio_frame.py

1import webrtcvad
2
3# Initialize the Voice Activity Detector
4vad = webrtcvad.Vad()
5
6# Set aggressiveness mode: 0 is least aggressive about filtering out non-speech, 3 is most aggressive.
7vad.set_mode(1)
8
9# Run the VAD on a 10ms piece of silence (160 samples at 16000Hz)
10# The frame must be 10, 20, or 30 ms long and the sample rate 8000, 16000, 32000 or 48000Hz.
11sample_rate = 16000
12frame_ms = 10
13frame = b'\x00\x00' * int(sample_rate * frame_ms / 1000)
14
15is_speech = vad.is_speech(frame, sample_rate)
16
17print(f"Is speech detected? {is_speech}")