Classifying Sounds Using a Neural Network

This video tutorial guides you through building a flexible sound classifier.

A Brief Synopsis

This tutorial guides you through building a sound classifier. It leverages a type of Neural Network in the Fluid Corpus Manipulation toolkit, the MLPClassifier to perform the classification process. This workflow is particularly flexible as you can train the MLPClassifier to associate any kind of data to a given label. In the context of this tutorial, we associate audio-descriptors to instrumental labels, so that an input sound can be classified according to a set of instrumental archetypes. There are lots of ways in which this workflow can be applied musically. For example you could:

  1. Drive audio-visual reactivity from the classification output
  2. Create an internal logic for an improvisational machine based on a sort of musical decision making
  3. Speed up the process of classifying a large corpus of sound resources.

Last modified: Tue Aug 23 14 by James Bradbury
Edit File on GitHub