{"ID":2898319,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2507.03395","arxiv_id":"2507.03395","title":"MaskBeat: Loopable Drum Beat Generation","abstract":"We present MaskBeat, a transformer-based approach for loopable drum pattern generation. Rather than predicting drum hits sequentially, our method uses bidirectional attention with iterative refinement, allowing instruments to be generated in parallel while maintaining musical coherence. Additionally, we introduce custom loss functions that capture drum-specific musical relationships. Our experiments show that MaskBeat generates higher quality and more musically coherent drum patterns than baseline approaches.","short_abstract":"We present MaskBeat, a transformer-based approach for loopable drum pattern generation. Rather than predicting drum hits sequentially, our method uses bidirectional attention with iterative refinement, allowing instruments to be generated in parallel while maintaining musical coherence. Additionally, we introduce custo...","url_abs":"https://arxiv.org/abs/2507.03395","url_pdf":"https://arxiv.org/pdf/2507.03395v1","authors":"[\"Luca A. Lanzendörfer\",\"Florian Grötschla\",\"Karim Galal\",\"Roger Wattenhofer\"]","published":"2025-07-04T08:57:24Z","proceeding":"cs.SD","tasks":"[\"cs.SD\",\"eess.AS\"]","methods":"[\"Transformer\"]","has_code":false}