Closed kangaro0 closed 2 years ago
Hey,
I'm trying out your package and I noticed that a recording only consisting of white noise gets detected as voiced. I'm cutting the whole recording in n frames (10ms, 16kHz, mono) and process them with a for-loop ( like in the example-file ).
Can you give me a hint on why this is happening?
King regards, kangaro0
Please give a try to silero-vad as well We posted a comparison here - https://github.com/wiseman/py-webrtcvad/issues/68
Hey,
I'm trying out your package and I noticed that a recording only consisting of white noise gets detected as voiced. I'm cutting the whole recording in n frames (10ms, 16kHz, mono) and process them with a for-loop ( like in the example-file ).
Can you give me a hint on why this is happening?
King regards, kangaro0