This workshop will teach how to handle sound files in python, compute sound and audio features from them, and run machine learning algorithms on them. This could be for anyone looking for supervised and unsupervised learning from sound files, using techniques for classification of sounds into one or more classes, and clustering of sounds based on extracted features. It will use Python 3 and some state of the art libraries in the field: Librosa, Pandas, and Scikit-Learn, all of which are open software, enabling people to be able to use their knowledge after the workshop.
I will provide installation instructions ahead of the workshop, and also have help pages online one week before the workshop. I will also be available to solve installation problems 1 hour before the start of the workshop.
After attending this workshop, learners are able to extract musical features from sound files, batch process sound files, and learn how to run supervised and unsupervised machine learning algorithms. This will not be a deep learning or a neural network focused workshop.
Some familiarity with python or programming is appreciated although not required. Some knowledge of audio features is also appreciated.
Please bring own laptop with python, LibROSA, pandas, numpy and scipy running.
Researchers working with audio data, for example, musicologists, linguists, media scientists; but others I haven't thought of such as physicists, chemists. It is important to note that I will not be covering the linguistic or phonological extraction aspects of speech corpora, this will be focused on audio analysis related tool kits.