Exploring Transformer-Based Music Overpainting for Jazz Piano Variations
This work addresses the challenge of limited scalability and diversity in music overpainting for jazz piano, though it appears incremental as it focuses on dataset scaling rather than novel methods.
This paper tackles the problem of generating jazz piano variations using transformer-based models for music overpainting, with results showing promising improvements in generalization and performance when trained on a larger dataset of 4,352 pairs compared to a smaller one.
This paper explores transformer-based models for music overpainting, focusing on jazz piano variations. Music overpainting generates new variations while preserving the melodic and harmonic structure of the input. Existing approaches are limited by small datasets, restricting scalability and diversity. We introduce VAR4000, a subset of a larger dataset for jazz piano performances, consisting of 4,352 training pairs. Using a semi-automatic pipeline, we evaluate two transformer configurations on VAR4000, comparing their performance with the smaller JAZZVAR dataset. Preliminary results show promising improvements in generalisation and performance with the larger dataset configuration, highlighting the potential of transformer models to scale effectively for music overpainting on larger and more diverse datasets.