Exploring Procedural Data Generation for Automatic Acoustic Guitar Fingerpicking Transcription

TL;DR

This paper presents a procedural data generation pipeline for training acoustic guitar transcription models, showing that synthetic data can effectively supplement real recordings and improve transcription accuracy.

Contribution

It introduces a novel procedural synthesis method for creating training data for guitar transcription, reducing reliance on scarce real recordings.

Findings

Synthetic data enables effective training of transcription models.

Finetuning with real data improves accuracy.

Procedural data can complement real recordings in music IR tasks.

Abstract

Automatic transcription of acoustic guitar fingerpicking performances remains a challenging task due to the scarcity of labeled training data and legal constraints connected with musical recordings. This work investigates a procedural data generation pipeline as an alternative to real audio recordings for training transcription models. Our approach synthesizes training data through four stages: knowledge-based fingerpicking tablature composition, MIDI performance rendering, physical modeling using an extended Karplus-Strong algorithm, and audio augmentation including reverb and distortion. We train and evaluate a CRNN-based note-tracking model on both real and synthetic datasets, demonstrating that procedural data can be used to achieve reasonable note-tracking results. Finetuning with a small amount of real data further enhances transcription accuracy, improving over models trained exclusively on real recordings. These results highlight the potential of procedurally generated audio for data-scarce music information retrieval tasks.