Document Type
Article
Publication Date
2-28-2026
Department
Department of Mechanical and Aerospace Engineering
Abstract
Sound quality attributes have been proven important predictors of customer satisfaction with consumer goods and appliances. The results of several studies indicate that level-based, tonal, and temporal aspects of sound influence the perceived quality of consumer products. The current state-of-the-art of inferring consumer satisfaction with sound attributes is based on the jury test methodology. However, product engineers often find the models generated from these studies incomplete, resulting in products that fail to meet consumer expectations. Therefore, this study aims to utilize generative data-driven approaches to create a range of acceptable sound quality attributes given consumer satisfaction requirements. A baseline sound obtained from the operational noise generated by a home appliance was used as a reference sound, while an orthogonally-designed experiment was conducted to prepare sound samples based on selected sound attributes. Consumer studies were conducted to train the data-driven models used in the proposed data-driven approach. Finally, a generative adversarial network (GAN) was utilized to generate an arrangement of the salient sound quality metrics to assist in developing a complete set of acoustic requirements to meet consumer satisfaction metrics. Therefore, the proposed methods are valuable and efficient as they eliminate the need for further listening-based jury studies in evaluating new products within the model inference space.
Publication Title
Noise Control Engineering Journal
Recommended Citation
Lesko, D. J.,
&
Nguyen, V.
(2026).
Informing Consumer Product Sound Quality Analysis using Generative Adversarial Networks.
Noise Control Engineering Journal,
74(1), 14-25.
http://doi.org/10.3397/1/37742
Retrieved from: https://digitalcommons.mtu.edu/michigantech-p2/2522
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 International License.
Version
Publisher's PDF
Publisher's Statement
© 2026. Publisher’s version of record: https://doi.org/10.3397/1/37742