AI Paper Atlas

Research map / Video / Video segmentation and tracking

Skipping the Frame-Level: Event-Based Piano Transcription With Neural Semi-CRFs

Yujia Yan, Frank Cwitkowitz, Zhiyao Duan

NeurIPS 2021 · Video › Video segmentation and tracking · approach: Event-based video reconstruction · 6 citations (OpenAlex)

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 changedWhereMeasuredEffect
Remove extra adjustable windows (oversampled log-mel) (ablation)Proposed semi-CRF · MAESTRO v2 note w/ offsetF1: 88.37 -> 88.63Improved
Remove final convolutional block in score matrix (ablation)Proposed semi-CRF · MAESTRO v2 note w/ offsetF1: 88.26 -> 88.63Improved
Remove interval-length scaling of scores (ablation)Proposed semi-CRF · MAESTRO v2 note w/ offsetF1: 88.51 -> 88.63Improved
Increase batch size from 12 to 20 (ablation)Proposed semi-CRF · MAESTRO v2 note w/ offset & velocityF1: 87.65 -> 87.75Improved
Split architecture into note and pedal branches sharing only conv blocks (ablation)Proposed semi-CRF · MAESTRO v2 pedal onsetF1: 78.10 -> 79.98Improved
Split architecture into note and pedal branches (ablation)Proposed semi-CRF · MAESTRO v2 pedal onset & offsetF1: 74.71 -> 76.85Improved
Use event-level semi-CRF instead of frame-level model with post-processingProposed vs Kong et al. · MAESTRO v2 note w/ offsetF1: 84.45 -> 88.63Improved
Use semi-CRF instead of frame-level modelProposed vs Kong et al. · MAESTRO v2 note onsetF1: 96.77 -> 96.39Hurt
Use semi-CRF instead of frame-level modelProposed vs Kong et al. · MAESTRO v2 pedal activationF1: 94.11 -> 90.73Hurt
Train on MAESTRO v3 instead of v2 splitsProposed (batchsize=20) · note w/ offset & velocityF1: 87.75 -> 87.44No clear effect
Use Proposed instead of Kong et al. for end-to-end transcription of a 33.3 minute fileProposed vs Kong et al. · Carl Czerny Grand Sonata Op.145 No.9running time (s): 353s -> 95sImproved
Use GPU instead of CPU for forward-backward on 1600 framessemi-CRF components · runtimeseconds: 7.12 -> 0.27Improved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy