Michigan Technological University
ASSESSING THE SOUND QUALITY OF CONSUMER PRODUCTS USING DATA-DRIVEN METHODOLOGIES
Abstract
dc:description.abstractProduct sound quality is an important predictor of customer satisfaction. Existing methodologies for consumer product sound quality are confined to enumerative approaches. While the current state-of-the-art can provide essential insights into the psychoacoustic performance of a product, it does not adequately address background variables, nor does it cover an inference space broad enough to assess the impact of interactions. This research applies systematic sound quality studies to assess the impact of significant background variables and interactions of level-based, tonal, and temporal acoustic attributes. These systematic studies resulted in a large set of consumer response data that was utilized to build and train data-driven models which can mathematically relate product psychoacoustics to consumer response. The results of this work show that background variables, such as product brand, can have a significant impact on the perception of a product’s sound quality. In addition, it was discovered that interactions of sound attributes are statistically significant factors in consumer product psychoacoustics. Finally, it was proven that data-driven methods, such as artificial neural networks and generative adversarial networks, can provide valuable insights into consumer product design and are significant contributions to advancing the state-of-the-art of consumer product sound quality.
Degree
thesis:*- Name thesis:degree_name
- Doctor of Philosophy in Mechanical Engineering-Engineering Mechanics (PhD)
- Level thesis:degree_level
- Campus Access Dissertation
- Discipline thesis:degree_discipline
- Department of Mechanical and Aerospace Engineering
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Lesko, Daniel Jeffery
- Committee members dc:contributor.committeemember
-
- Jason Blough
- Susanta Ghosh
- Shane Mueller
- Contributors dc:contributor
-
- Vinh Nguyen
Subjects
dc:subject × 6Identifiers
dc:identifier.*- Identifier
- https://digitalcommons.mtu.edu/etdr/1854