Audio Classification model


I’m trying to make the model in Himax WE-1 board, using built-in microphone.

I’m wonder edge-impulse audio model can classify pitch of musical instruments.
Our team planned to build the model using instruments (recorded on Himax WE-1 board) to classify “sol”, “do”. Is it possible to distinguish??

Another curious thing is, I would like to get a document with an explanation for each impulse. For example, how do I adjust fft in MFCC and I would like to get this information.

Best Regards.

Hi @SunBeenMoon Yeah that should be possible, I’d go with a normal spectrogram block. Here’s the processing blocks docs:

Thank you for your reply @janjongboom !!
I create awesome project!!

project number is 22544

1 Like