Beginner Audio Classification

Can you recognize multiple sounds within a single model? I followed the tutorial and example for classifying the sound of a faucet (running water from a tap). Could you adapt this model to train it to recognize two distinct sounds? For example, Sound 1 might detect water flowing from a faucet into a sink, while Sound 2 might detect water from an electric shower. The model I want to build for learning purposes is designed to detect whether a plucked guitar string is a low E, a low A, or Unknown.

I plan to use a raspberry PI for this,

Hi @spc_ml,

Yes, you can have a model recognize multiple outputs. You just need to set the labels, and Edge Impulse will do the rest to configure the model for your. For example, the faucet example likely has two labels: faucet and background. You’d want to have 3 (or more) labels, such as faucet, shower, background.

1 Like

Thanks for the response