The goal of this project is to explore and evaluate ways of generating sheet music images. We do this by training two language models, AWD-LSTM and GPT-2, and by using 5 different approaches: pixel columns, image patches (pixel blocks), visual word tokens, semantic encoding, and XML/MEI code.
This work was presented at ISMIR 2022. The paper can be found here.
Marcos Acosta, Irmak Bukey, and TJ Tsai. "An Exploration of Generating Sheet Music Images." Proceedings of the International Society for Music Information Retrieval Conference (ISMIR), 2022, pp. 701-708.
This material is based upon work supported by the National Science Foundation under Grant No. 2144050. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.