Salvato in:
Dettagli Bibliografici
Autori principali: Bourdin, Yann, Legrand, Pierrick, Roche, Fanny
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:https://arxiv.org/abs/2512.07393
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866912753917100032
author Bourdin, Yann
Legrand, Pierrick
Roche, Fanny
author_facet Bourdin, Yann
Legrand, Pierrick
Roche, Fanny
contents This paper investigates the optimization of Truncated Backpropagation Through Time (TBPTT) for training neural networks in digital audio effect modeling, with a focus on dynamic range compression. The study evaluates key TBPTT hyperparameters -- sequence number, batch size, and sequence length -- and their influence on model performance. Using a convolutional-recurrent architecture, we conduct extensive experiments across datasets with and without conditionning by user controls. Results demonstrate that carefully tuning these parameters enhances model accuracy and training stability, while also reducing computational demands. Objective evaluations confirm improved performance with optimized settings, while subjective listening tests indicate that the revised TBPTT configuration maintains high perceptual quality.
format Preprint
id arxiv_https___arxiv_org_abs_2512_07393
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Empirical Results for Adjusting Truncated Backpropagation Through Time while Training Neural Audio Effects
Bourdin, Yann
Legrand, Pierrick
Roche, Fanny
Machine Learning
This paper investigates the optimization of Truncated Backpropagation Through Time (TBPTT) for training neural networks in digital audio effect modeling, with a focus on dynamic range compression. The study evaluates key TBPTT hyperparameters -- sequence number, batch size, and sequence length -- and their influence on model performance. Using a convolutional-recurrent architecture, we conduct extensive experiments across datasets with and without conditionning by user controls. Results demonstrate that carefully tuning these parameters enhances model accuracy and training stability, while also reducing computational demands. Objective evaluations confirm improved performance with optimized settings, while subjective listening tests indicate that the revised TBPTT configuration maintains high perceptual quality.
title Empirical Results for Adjusting Truncated Backpropagation Through Time while Training Neural Audio Effects
topic Machine Learning
url https://arxiv.org/abs/2512.07393