Exploring Quality and Generalizability in Parameterized Neural Audio Effects
×
Cite This
Citation & Abstract
W. Mitchell, and SC. H.. Hawley, "Exploring Quality and Generalizability in Parameterized Neural Audio Effects," Paper 10397, (2020 October.). doi:
W. Mitchell, and SC. H.. Hawley, "Exploring Quality and Generalizability in Parameterized Neural Audio Effects," Paper 10397, (2020 October.). doi:
Abstract: This work expands on prior research published [1] on modeling nonlinear time-dependent signal processing effects by means of a deep neural network with parameterized controls, with the goal of producing commercially viable, high quality audio, i.e. 44.1kHz sampling rate at 16-bit resolution. These results highlight progress in modeling these effects through architecture and optimization changes, towards increasing computational efficiency, lowering signal-to-noise ratio, and extending to a larger variety of nonlinear audio effects. Most of the presented methods provide marginal or no increase in output accuracy over the original model, with the exception of dataset manipulation. We found that limiting the audio content of the dataset provided a significant improvement in model accuracy over models trained on more general datasets.
@article{mitchell2020exploring,
author={mitchell, william and hawley, scott h.},
journal={journal of the audio engineering society},
title={exploring quality and generalizability in parameterized neural audio effects},
year={2020},
volume={},
number={},
pages={},
doi={},
month={october},}
@article{mitchell2020exploring,
author={mitchell, william and hawley, scott h.},
journal={journal of the audio engineering society},
title={exploring quality and generalizability in parameterized neural audio effects},
year={2020},
volume={},
number={},
pages={},
doi={},
month={october},
abstract={this work expands on prior research published [1] on modeling nonlinear time-dependent signal processing effects by means of a deep neural network with parameterized controls, with the goal of producing commercially viable, high quality audio, i.e. 44.1khz sampling rate at 16-bit resolution. these results highlight progress in modeling these effects through architecture and optimization changes, towards increasing computational efficiency, lowering signal-to-noise ratio, and extending to a larger variety of nonlinear audio effects. most of the presented methods provide marginal or no increase in output accuracy over the original model, with the exception of dataset manipulation. we found that limiting the audio content of the dataset provided a significant improvement in model accuracy over models trained on more general datasets.},}
TY - paper
TI - Exploring Quality and Generalizability in Parameterized Neural Audio Effects
SP -
EP -
AU - Mitchell, William
AU - Hawley, Scott H.
PY - 2020
JO - Journal of the Audio Engineering Society
IS -
VO -
VL -
Y1 - October 2020
TY - paper
TI - Exploring Quality and Generalizability in Parameterized Neural Audio Effects
SP -
EP -
AU - Mitchell, William
AU - Hawley, Scott H.
PY - 2020
JO - Journal of the Audio Engineering Society
IS -
VO -
VL -
Y1 - October 2020
AB - This work expands on prior research published [1] on modeling nonlinear time-dependent signal processing effects by means of a deep neural network with parameterized controls, with the goal of producing commercially viable, high quality audio, i.e. 44.1kHz sampling rate at 16-bit resolution. These results highlight progress in modeling these effects through architecture and optimization changes, towards increasing computational efficiency, lowering signal-to-noise ratio, and extending to a larger variety of nonlinear audio effects. Most of the presented methods provide marginal or no increase in output accuracy over the original model, with the exception of dataset manipulation. We found that limiting the audio content of the dataset provided a significant improvement in model accuracy over models trained on more general datasets.
This work expands on prior research published [1] on modeling nonlinear time-dependent signal processing effects by means of a deep neural network with parameterized controls, with the goal of producing commercially viable, high quality audio, i.e. 44.1kHz sampling rate at 16-bit resolution. These results highlight progress in modeling these effects through architecture and optimization changes, towards increasing computational efficiency, lowering signal-to-noise ratio, and extending to a larger variety of nonlinear audio effects. Most of the presented methods provide marginal or no increase in output accuracy over the original model, with the exception of dataset manipulation. We found that limiting the audio content of the dataset provided a significant improvement in model accuracy over models trained on more general datasets.
Authors:
Mitchell, William; Hawley, Scott H.
Affiliation:
Belmont University, Nashville, TN, USA
AES Convention:
149 (October 2020)
Paper Number:
10397
Publication Date:
October 22, 2020Import into BibTeX
Subject:
Audio Processing
Permalink:
http://www.aes.org/e-lib/browse.cfm?elib=20934