Recently, after co-chairing the session on Machine Learning for Musical Acoustics along with the wonderful Bozena Kostek at the Acoustical Society of America Meeting in San Diego, a student came up to me to ask for advice or resources on how to start learning machine learning for audio applications.

And I didn’t have a good answer. Because, while there’s plenty of tips out there for learning image classification or NLP, not a ton of it relates to audio. Students have asked me this before…and so I was frustrated by my own lack of an answer, apart from “read whatever you find, try out demos, take whatever courses you can, etc.”

SO! Now, with graduate student Ryan Miller (EDIT 6/1/2020: who is now gainfully employed at Creative Labs!), we compiled a list and posted it to GitHub:

ml-audio-start Page

… and soliciting contributions from the community.

EDIT 02/01/2020: We now have contributions from accomplished researchers at Google, Adobe, and others!


(c) 2020 Scott H. Hawley