MaskBeat: Loopable Drum Beat Generation

2025/07/04 by Lanzendörfer, Luca A., Grötschla, Florian, Galal, Karim +1
#Audio and Speech Processing (eess.AS) #FOS: Computer and information sciences #FOS: Electrical engineering #Sound (cs.SD) #electronic engineering #information engineering

paper · doi:10.48550/arxiv.2507.03395

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.

Related

• Setting the rhythm scene: deep learning-based drum loop generation from arbitrary language cues
• Generating Coherent Drum Accompaniment With Fills And Improvisations
• The Rhythm In Anything: Audio-Prompted Drums Generation with Masked Language Modeling
• JukeDrummer: Conditional Beat-aware Audio-domain Drum Accompaniment Generation via Transformer VQ-VAE
• High-Level Control of Drum Track Generation Using Learned Patterns of Rhythmic Interaction