Audio Foundation Models Outperform Symbolic Representations for Piano Performance Evaluation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Author: Dhiman, Jai
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911401717530624
author Dhiman, Jai
author_facet Dhiman, Jai
contents Automated piano performance evaluation traditionally relies on symbolic (MIDI) representations, which capture note-level information but miss the acoustic nuances that characterize expressive playing. I propose using pre-trained audio foundation models, specifically MuQ and MERT, to predict 19 perceptual dimensions of piano performance quality. Using synthesized audio from PercePiano MIDI files (rendered via Pianoteq), I compare audio and symbolic approaches under controlled conditions where both derive from identical source data. The best model, MuQ layers 9-12 with Pianoteq soundfont augmentation, achieves R^2 = 0.537 (95% CI: [0.465, 0.575]), representing a 55% improvement over the symbolic baseline (R^2 = 0.347). Statistical analysis confirms significance (p < 10^-25) with audio outperforming symbolic on all 19 dimensions. I validate the approach through cross-soundfont generalization (R^2 = 0.534 +/- 0.075), difficulty correlation with an external dataset (rho = 0.623), and multi-performer consistency analysis. Analysis of audio-symbolic fusion reveals high error correlation (r = 0.738), explaining why fusion provides minimal benefit: audio representations alone are sufficient. I release the complete training pipeline, pretrained models, and inference code.
format Preprint
id arxiv_https___arxiv_org_abs_2601_19029
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Audio Foundation Models Outperform Symbolic Representations for Piano Performance Evaluation
Dhiman, Jai
Sound
Audio and Speech Processing
I.2.6; H.5.5; J.5
Automated piano performance evaluation traditionally relies on symbolic (MIDI) representations, which capture note-level information but miss the acoustic nuances that characterize expressive playing. I propose using pre-trained audio foundation models, specifically MuQ and MERT, to predict 19 perceptual dimensions of piano performance quality. Using synthesized audio from PercePiano MIDI files (rendered via Pianoteq), I compare audio and symbolic approaches under controlled conditions where both derive from identical source data. The best model, MuQ layers 9-12 with Pianoteq soundfont augmentation, achieves R^2 = 0.537 (95% CI: [0.465, 0.575]), representing a 55% improvement over the symbolic baseline (R^2 = 0.347). Statistical analysis confirms significance (p < 10^-25) with audio outperforming symbolic on all 19 dimensions. I validate the approach through cross-soundfont generalization (R^2 = 0.534 +/- 0.075), difficulty correlation with an external dataset (rho = 0.623), and multi-performer consistency analysis. Analysis of audio-symbolic fusion reveals high error correlation (r = 0.738), explaining why fusion provides minimal benefit: audio representations alone are sufficient. I release the complete training pipeline, pretrained models, and inference code.
title Audio Foundation Models Outperform Symbolic Representations for Piano Performance Evaluation
topic Sound
Audio and Speech Processing
I.2.6; H.5.5; J.5
url https://arxiv.org/abs/2601.19029