Contribution
Towards a binaural audio quality model for listeners with normal and impaired hearing
* Presenting author
Abstract:
Recently, Eurich et al. (2024) suggested the computationally efficient monaural and binaural audio quality model (eMoBi-Q) for normal-hearing listeners. eMoBi-Q combines monaural and binaural features and was successfully evaluated with six audio databases including quality ratings for music and speech processed by loudspeakers and algorithms typically applied in modern hearing devices (e.g., acoustic transparency, feedback cancellation or binaural beamforming). Here, we extend eMoBi-Q towards predicting perceptual consequences of hearing loss (HL) on audio quality perception. Given that unsatisfactory loudness perception is one of the most common complaints among hearing aid users, we aim to integrate loudness as a sub-measure for the prediction of audio quality for listeners with normal and impaired hearing. While predicting loudness itself is important in the context of loudness-based hearing aid fitting, loudness as audio quality sub-measure may be helpful for the selection of reliable auditory features in hearing impaired listeners. Thus, eMoBi-Q was extended by a nonlinear filterbank to account for effects of cochlear HL. Parameters of the filterbank and subsequent stages were motivated by the physiologically-motivated (binaural) loudness model of Pieper et al. (2018). Here a first version of the extended eMoBi-Q will be presented and discussed.