Research map / Image generation / Autoregressive and latent generation
Bridging Piano Transcription and Rendering via Disentangled Score Content and Style
Wei Zeng, JUNCHUAN ZHAO, Ye Wang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| train jointly with APT and unpaired data (vs EPR-only) (ablation) | EPR model · ASAP EPR | velocity MAE: 10.65 -> 10.33 | Improved |
| train jointly with APT and unpaired data (vs EPR-only) (ablation) | EPR model · ASAP EPR | velocity KL divergence: 2.05 -> 1.76 | Improved |
| use PSR-generated style instead of extracted target style | Joint model · ASAP EPR | velocity MAE: 10.33 -> 15.24 | Hurt |
| use PSR-generated style instead of extracted target style | Joint model · ASAP EPR | alignment rate (%): 91.55 -> 92.27 | Improved |
| use joint seq2seq model instead of end-to-end APT baseline | End-to-end APT · ASAP APT (MUSTER) | Eavg (%): 14.10 -> 12.48 | Improved |
| use joint seq2seq model instead of end-to-end APT baseline | End-to-end APT · ASAP APT (MUSTER) | pitch edit distance Ep (%): 2.73 -> 3.08 | Hurt |
| use joint model instead of end-to-end baseline for spelling | End-to-end APT · ASAP APT (ScoreSimilarity) | Espell (%): 14.31 -> 6.24 | Improved |
| use style embedding instead of score content for composer ID | Content representation · ATEPP composer ID | accuracy (%): 29.99 -> 77.46 | Improved |
| use style embedding instead of score content for performer ID | Content representation · ATEPP performer ID | accuracy (%): 9.94 -> 42.07 | Improved |
| use alignment-free joint model vs note-aligned APT baselines | Liu 2022 / Shibata 2021 · ASAP APT (MUSTER) | pitch edit distance Ep (%): 2.02 and 2.09-2.11 vs 3.08 | Hurt |
Closest papers by meaning
- Aria-MIDI: A Dataset of Piano MIDI Files for Symbolic Music Modeling (ICLR 2025 · similarity 0.86)
- Rubato: Transcribing Piano Music with Timestamps (NeurIPS 2026 · similarity 0.86)
- Structured Multi-Track Accompaniment Arrangement via Style Prior Modelling (NeurIPS 2024 · similarity 0.85)
- MIDI-GPT: A Controllable Generative Model for Computer-Assisted Multitrack Music Composition (AAAI 2025 · similarity 0.84)
- Unifying Symbolic Music Arrangement: Track-Aware Reconstruction and Structured Tokenization (NeurIPS 2025 · similarity 0.84)
- MusER: Musical Element-Based Regularization for Generating Symbolic Music with Emotion (AAAI 2024 · similarity 0.84)
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.
