Diffroll: Diffusion-based Generative Music Transcription With Unsupervised Pretraining Capability

Abstract

In this paper we propose a novel generative approach, DiffRoll, to tackle automatic music transcription (AMT). Instead of treating AMT as a discriminative task in which the model is trained to convert spectrograms into piano rolls, we think of it as a conditional generative task where we train our model to generate realistic looking piano rolls from pure Gaussian noise conditioned on spectrograms. This new AMT formulation enables DiffRoll to transcribe, generate and even inpaint music. Due to the classifier-free nature, DiffRoll is also able to be trained on unpaired datasets where only piano rolls are available. Our experiments show that DiffRoll outperforms its discriminative counterpart by 19 percentage points (ppt.) and our ablation studies also indicate that it outperforms similar existing methods by 4.8 ppt. Source code and demonstration are available https://sony.github.io/DiffRoll/.

Stats

• citations12
• S2 citations—
• github stars0
• HF likes0
• heat score8.35
• arxiv keycheuk2022diffroll

Related papers

• D3RM: A Discrete Denoising Diffusion Refinement Model For Piano Transcription (2025)5.93
• Piano Transcription By Hierarchical Language Modeling With Pretrained Roll-based Encoders (2025)4.52
• Diff-a-riff: Musical Accompaniment Co-creation Via Latent Diffusion Models (2024)0.00
• Annotation-free Automatic Music Transcription With Scalable Synthetic Data And Adversarial Domain Confusion (2023)4.52
• Diffrhythm+: Controllable And Flexible Full-length Song Generation With Preference Optimization (2025)3.58
• Motionrag-diff: A Retrieval-augmented Diffusion Framework For Long-term Music-to-dance Generation (2025)0.00
• Diffrhythm: Blazingly Fast And Embarrassingly Simple End-to-end Full-length Song Generation With Latent Diffusion (2025)0.00
• Musicldm: Enhancing Novelty In Text-to-music Generation Using Beat-synchronous Mixup Strategies (2023)13.55