Research map / Graph learning / Knowledge and temporal graphs
Graph Neural Network for Music Score Data and Modeling Expressive Piano Performance
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use ISGN graph encoder instead of note-wise LSTM baseline | expressive performance VAE · Yamaha e-competition test set | tempo MSE: 0.2721 -> 0.2379 | Helped |
| use ISGN instead of baseline | expressive performance VAE · Yamaha e-competition test set | velocity MSE: 0.6011 -> 0.5877 | Helped |
| use ISGN instead of baseline | expressive performance VAE · Yamaha e-competition test set | pedal MSE: 0.8056 -> 0.7544 | Helped |
| use ISGN instead of baseline | expressive performance VAE · Yamaha e-competition test set | onset deviation MSE: 0.7678 -> 0.7978 | Hurt |
| use ISGN instead of baseline | expressive performance VAE · Yamaha e-competition test set | KLD: 2.2581 -> 3.7247 | Hurt |
| use ISGN instead of HAN | expressive performance VAE · Yamaha e-competition test set | KLD: 13.666 -> 3.7247 | Helped |
| replace note/voice LSTM with GGNN (G-HAN) vs HAN | HAN · Yamaha e-competition test set | tempo MSE: 0.2380 -> 0.2785 | Hurt |
| add HAN hierarchy to baseline note LSTM | BL · Yamaha e-competition test set | tempo MSE: 0.2721 -> 0.2380 | Helped |
| use ISGN instead of baseline | expressive performance VAE · Tempo correlation, 50 excerpts | excerpts with r > 0.7: 29 -> 35 | Helped |
| use ISGN instead of baseline | expressive performance VAE · Tempo correlation, 50 excerpts | excerpts with r > 0.5: 36 -> 42 | Helped |
| use ISGN instead of G-HAN | expressive performance VAE · Tempo correlation, 50 excerpts | average max r: 0.6994 -> 0.6821 | No clear effect |
| use ISGN instead of G-HAN | expressive performance VAE · Tempo correlation, 50 excerpts | excerpts with r > 0.9: 21 -> 18 | Hurt |
Closest papers by meaning
- PerformanceNet: Score-to-Audio Music Generation with Multi-Band Convolutional Residual Network (AAAI 2019 · similarity 0.89)
- AutoSchA: Automatic Hierarchical Music Representations via Multi-Relational Node Isolation (AAAI 2026 · similarity 0.88)
- N-gram Unsupervised Compoundation and Feature Injection for Better Symbolic Music Understanding (AAAI 2024 · similarity 0.84)
- Aria-MIDI: A Dataset of Piano MIDI Files for Symbolic Music Modeling (ICLR 2025 · similarity 0.84)
- Rubato: Transcribing Piano Music with Timestamps (NeurIPS 2026 · similarity 0.83)
- Structured Multi-Track Accompaniment Arrangement via Style Prior Modelling (NeurIPS 2024 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.