Musical Gesture Recognition Using Machine Learning and Audio Descriptors

We report preliminary results of an ongoing project on automatic recognition and classification of musical “gestures” from audio extracts. We use a machine learning tool designed for motion tracking and recognition, applied to labeled vectors of audio descriptors in order to recognize hypothetical gestures formed by these descriptors. A hypothesis is that the classes detected in audio descriptors can be used to identify higher-level/abstract musical structures which might not be described easily using standard/symbolic representations.