Machine Learning for acoustics

Hi, I was experimenting with machine learning tools to identify certain sound behavior of engines and gear systems. I used my own set of scripts to extract features (3rd octave, sound pressure level, ..) of my datasets.

For image and speech recognition there are already preprocessing methods and pretrained Convolutional Neural Networks. I wonder if there is anything comparable for general acoustic issues (not speech recognition).

Could anyone provide links to resources or whitepapers on this subject?

Answer 1

[Niraj Gadakari]

Here is a link to an article about applying Machine Learning techniques to classify musical instrument loudspeakers. The article discusses in detail about the entire workflow, from creating labelled data, selecting features to train machine learning algorithms, applying algorithms, and fine tuning the learned model.