Ask HN: Audio analysis models, how to train to learn sound patters?

4 pointsby thedanglervor 6 Monaten2 comments

Hello all, I'm looking to for sound models that understand sound patters that can be used locally.

Hoping I can use them on device like iOS CoreML.

Before I get started it doesn't hurt to ask to see if anyone has done this before or has a sound model that detects patters, Anger, excitement, Arousal, laughing, crying, laughing, human emotions in general.

I found this :https://medium.com/@narner/classification-of-sound-files-on-...

But maybe there is a better way or something already exists.

Thanks

2 comments

atoavvor 6 Monaten

Python has a library called librosa that helps with that. What you are looking for is called "Feature Extraction". Which kind of extraction is the best for you depends on what you are trying to extract (percussive sounds, speech, instruments etc)

thedanglervor 6 Monaten

Thank you! I'll give it a shot.