Audio recognition

I want to start an audio recognition project on the jetson nano.

  • How easy/hard is it to realise a microphone as inputdevice for the jetson nano? Is there already a file where i just have to change some settings like the just for audio or can i just add a parser argument like for the camera?
parser.add_argument("--camera", type=str, default="/dev/video0")
  • Is it possible to use a microphone embedded in the camera to use as input device as well or do i take an extra microphone with additional usb port?

I would be very thankful if you can help me out with that or have an additional tutorial for me to follow for the beginning


Hi steffen.epp,

You could refer to below thread to see if can help:

Thank you for your fast answer, the first link is written by myself and is about something else.
The second link is about adding a speaker to the nano and the software to make it able to speak, but not so much about voice recognition.

I already got my Model running and ready to use. In my case, I want to run a python code on the nano where I use a microphone as input to categorize spoken words. But I don’t know how I can set my microphone as my inputdevice to get data from which i want to categorize


Here are some information about USB-based microphone for your reference first:

This command shared in #1 is actually using GStreamer interface within a python wrapper.
GStreamer also provide an audio interface, which can be found in the following document:


Hi @AastaLLL
I have a logitech usb c525 webcamera on my jetson with built in microphone. I wont to use the audio input from the camera for audio recognition. When checking the sound settings, the microphone seems to work. But when i try the gstream command it only makes a file with a beeping sound, which are much longer than I recorded for. Any addvice?

Hi vegard-si,

Please open a new topic for your issue. Thanks