# SCISSOR: Score-Conditioned Instrument Source Separation for Orchestral Recordings

arXiv: https://arxiv.org/abs/2609.33265 · HTML: https://mlsanity.com/paper/2609.33265/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 10 of its 16 references that are in the map.
- First posted on arXiv 27 Sep 2026.

## Builds on (most cited first, 10 of 10)

- [Hybrid Transformers for Music Source Separation](https://mlsanity.com/paper/2211.08553/) (2022)
- [Music Source Separation with Band-split RNN](https://mlsanity.com/paper/2209.15174/) (2022)
- [Creating A Multi-track Classical Musical Performance Dataset for Multimodal Music Analysis: Challenges, Insights, and Applications](https://mlsanity.com/paper/1612.08727/) (2016)
- [The 2018 Signal Separation Evaluation Campaign](https://mlsanity.com/paper/1804.06267/) (2018)
- [Music Source Separation with Band-Split RoPE Transformer](https://mlsanity.com/paper/2309.02612/) (2023)
- [All for One and One for All: Improving Music Separation by Bridging Networks](https://mlsanity.com/paper/2010.04228/) (2020)
- [Structured Dropout for Weak Label and Multi-Instance Learning and Its Application to Score-Informed Source Separation](https://mlsanity.com/paper/1609.04557/) (2016)
- [An Overview of Lead and Accompaniment Separation in Music](https://mlsanity.com/paper/1804.08300/) (2018)
- [SynthSOD: Developing an Heterogeneous Dataset for Orchestra Music Source Separation](https://mlsanity.com/paper/2409.10995/) (2024)
- [Score-informed Music Source Separation: Improving Synthetic-to-real Generalization in Classical Music](https://mlsanity.com/paper/2503.07352/) (2025)

## Abstract (arXiv)

Orchestral separation recovers instrument sections from mixtures in which shared pitches, harmonics, and timbres obscure source identity. An aligned score provides instrument labels, note pitches, and activity times. A score-informed approach appends piano rolls to audio features before mask prediction. We introduce SCISSOR (Score-Conditioned Instrument Source Separation for Orchestral Recordings), which uses the score to form a frame-wise query for each source. Each query matches a shared audio representation, and a softmax over instrument and background slots jointly assigns overlapping time-frequency evidence. The queries retain instrument identity even when notes are missing from the score. After training on SynthSOD and a small set of URMP and PHENICX-Anechoic recordings, SCISSOR achieves the highest average SDR on held-out real recordings. With SynthSOD-only training, it leads on SynthSOD and zero-shot PHENICX-Anechoic, and improves on its audio-only control on zero-shot URMP. SCISSOR also degrades less under score corruption than the evaluated score-based baselines.