publications

2024

1. Advancing AI in Music Composition: Refining the Generative Music Overpainting Task

   E. Row, and G. Fazekas

   In AES International Symposium on AI and the Musician, Apr 2024

   Abs PDF

   Recent advancements in Deep Learning (DL) have greatly improved the outputs of music generation tasks. Amidst the debate on the impact of Artificial Intelligence (AI) on music creation within the music industry and concerns over artistic control, our paper proposes utilising smaller generative music tasks based on existing music composition techniques. We expand upon the task of Music Overpainting, which aims to generate variations on small music sections. The Music Overpainting task, was initially explored using the JAZZVAR dataset - a collection of original and variation MIDI segments from jazz standards. Using this dataset, we present a more robust transformer base- line, allowing for improved benchmarking for the task and discuss some approaches that will contribute to a comprehensive evaluation framework for these types of tasks.

2023

1. JAZZVAR: A Dataset of Variations found within Solo Piano Performances of Jazz Standards for Music Overpainting

   Eleanor Row, Jingjing Tang, and George Fazekas

   In Proceedings of the 16th International Symposium on Computer Music Multidisciplinary Research, Jul 2023

   Abs arXiv PDF Listening Examples Video Presentation

   Jazz pianists often uniquely interpret jazz standards. Passages from these interpretations can be viewed as sections of variation. We manually extracted such variations from solo jazz piano performances. The JAZZVAR dataset is a collection of 502 pairs of Variation and Original MIDI segments. Each Variation in the dataset is accompanied by a corresponding Original segment containing the melody and chords from the original jazz standard. Our approach differs from many existing jazz datasets in the music information retrieval (MIR) community, which often focus on improvisation sections within jazz performances. In this paper, we outline the curation process for obtaining and sorting the repertoire, the pipeline for creating the Original and Variation pairs, and our analysis of the dataset. We also introduce a new generative music task, Music Overpainting, and present a baseline Transformer model trained on the JAZZVAR dataset for this task. Other potential applications of our dataset include expressive performance analysis and performer identification.