Explainability in music deep learning models
Organization
Computational music analysis and recommendation
Abstract
Recently, autoregressive language models (transformers) have become SOTA in speech and music domain as well by using tricks to diminish the size of sequence while keeping a large receptive field. We would benefit from understanding what is the musical meaning behind learned tokens.
Graduation Theses defence year
2022-2023
Supervisor
Anna Aljanaki
Spoken language (s)
Estonian, English
Requirements for candidates
Level
Masters
Keywords
Application of contact
Name
Anna Aljanaki
Phone
E-mail