MR-MT3: Memory Retaining Multi-Track Music Transcription to Mitigate Instrument Leakage

arXiv:2403.10024v1 Announce Type: cross
Abstract: This paper presents enhancements to the MT3 model, a state-of-the-art (SOTA) token-based multi-instrument automatic music transcription (AMT) model. Despite SOTA performance, MT3 has the issue of instrument leakage, where transcriptions are fragmented across different instruments. To mitigate this, we propose MR-MT3, with enhancements including a memory retention mechanism, prior token sampling, and token shuffling are proposed. These methods are evaluated on the Slakh2100 dataset, demonstrating improved onset F1 scores and reduced instrument leakage. In addition …

abstract art arxiv cs.ai cs.lg cs.mm cs.sd eess.as issue memory music paper performance sota state token transcription type