This study investigates how tonality influences the perceived annoyance of heat pump noise. Heat pumps produce noise mainly from fans and compressors. These noises often contain prominent tonal components that can make them particularly disturbing. In this work, four fundamental heat pump noise samples were divided into three frequency bands, and their tonal features were enhanced to generate transformed noise samples. A jury test was conducted using a semantic differential listening experiment, in which rated the noises were rated on the basis of ten descriptive adjective pairs. Statistical analyses revealed that noises with higher frequency and higher tonality were significantly more annoying, whereas low tonality noise had a relatively minor impact. Additionally, a deep transfer learning model based on MobileNetV2 was trained on the spectrograms to classify them as „annoying“ or „not annoying,“ achieving an accuracy of 78%. Overall, the findings indicate that both frequency and tonality are key factors in noise annoyance, and deep transfer learning methods can effectively predict this annoyance.

Melden Sie sich hier an um Einladungen zu den Kolloquium-Vorträgen per E-Mail zu erhalten.
Register here to receive the invitations to colloquium talks via e-mail.

Zoom-Meeting-ID: 954 4073 3814
Passwort: 450783