ReSpeaker Mic Array v2.0 and Speech recognition configuration

Products & Technology ReSpeaker
#1

I need to use a high sensitivity microphone on a device that involves speech recognition. I selected Mic Array v2.0 and i had problems on speech recognition. After searching to find the cause, i rea on google docs for speech recognition the following:

  • Do not use automatic gain control (AGC).
  • All noise reduction processing should be disabled.
    Both of them are enabled on default on Mic Array v2.0. How can i disable them ?
    Thank you in advance
#2

Checkout the GitHub project: https://github.com/respeaker/usb_4_mic_array

Prepare Python venv:

python3 -m venv venv
source venv/bin/activate
pip3 install wheel
pip3 -r requirements.txt

List all parameters:

python tuning.py -r

Set parameter:

python tuning.py AGCONOFF 0

Search for other parameters to set.

1 Like
#3

I have done that. Thank you very much. Do you know if i should dissable “echo cancelation” for speech recognition?

#4

No, but echo cancellation sounds actually very useful. Best is to use a recording software like audacity, record all six channel and try out what works best.

#5

Hi, May I know how I can configure the ReSpeaker Mic Array v2.0 to beamform at a particular direction say 60 deg? or 90 deg? Seems to have no documentation on this part?
Thanks