Deep Learning for Sound Recognition

Leveraging the breadth of expertise at UAlberta in sound studies and machine learning, this cluster aims to develop supervised machine learning algorithms capable of automatically annotating audio recordings. These algorithms will allow audio assets to be consistently annotated across multiple repositories with reasonable accuracy, and the potential to revolutionize disciplines that depend on searching, indexing, classifying, and analyzing audio collections, including ethnomusicology, linguistics, and bioacoustics.

Principal Investigator

Michael Frishkopf (Music/Canadian Centre for Ethnomusicology)