Machine learning in audio?

Question

What are recommended books or online courses that are mostly helpful to someone wanting to delve into machine learning specifically in audio?

Or is the learning path mostly the same as for any other machine learning fields?

Answer 1

I think that resources about speech processing would be a good introduction to audio files manipulation for machine learning in general.

For a broader introduction to audio, Machine Learning for Audio, Image and Video Analysis: Theory and Applications, Reviews Book by Alessandro Vinciarelli and Francesco Camastra could be a good read.

Also, check out some blogs like this one: https://www.analyticsvidhya.com/blog/2018/01/10-audio-processing-projects-applications/ (as you mention deep learning in the tags) that can give a good introduction with some hands-on experience.

Answer 2

There is now a youtube channel dedicated to Audio Machine Learning, The Sound of AI. There is also an associated Slack community for discussions.

Answer 3

Using Deep learning in audio processing is a very interesting field of study and there are not a large number of companies actually working on this problem set, so finding the relevant material is a little hard for this. But if you want to refer to the development and understand the thinking behind processing audio in images, refer to these links:

• Looking to Listen: Audio-Visual Speech Separation
• Representations of Sound in Deep Learning of Audio Features from Music
• Unsupervised feature learning for audio classification

Answer 4

For general sound, I recommend Computational Analysis of Sound Scenes and Events. Music and speech are popular sub-fields of audio that have a lot of literature dedicated to them, even before machine learning became a thing.