Speech quality and intelligibility are of significant importance during clinical hearing aid (HA) fitting and verification. Validated intrusive objective predictors of intelligibility and quality such as the Hearing Aid Speech Perception Index (HASPI) and the Hearing Aid Speech Quality Index (HASQI) have not been widely adopted for implementation within clinically available HA test systems. Recent advances in non-intrusive measures, such as those from Clarity Prediction Challenges (CPCs) and HASA-Net, are also not yet accessible to clinicians. Moreover, most of these advancements rely on datasets from simulated HAs, not the commercial devices used by audiologists. This work aims to develop non-intrusive quality and intelligibility indices using custom databases of noisy speech recorded in a HA test box. Similar to the successful models from previous CPCs, the indices were derived using a novel non-intrusive model leveraging automatic speech recognition and self-supervised learning techniques. The proposed non-intrusive model was trained to predict the intrusive HASPI/HASQI values and later validated against subjective intelligibility data obtained from a group of listeners with hearing loss. The proposed model resulted in strong correlations with HASQI (95.96%) and HASPI (97.59%), and a moderate correlation with the subjective intelligibility scores (76.63%).

Objective measures of speech quality are highly desirable in benchmarking and monitoring the performance of hearing aids (HAs). Existing HA speech quality indices such as the hearing aid speech quality index (HASQI) are intrusive in that they require a properly time-aligned and frequency-shaped reference signal to predict the quality of HA output. Two new reference-free HA speech quality indices are proposed in this paper, based on a model that amalgamates perceptual linear prediction (PLP), hearing loss (HL) modeling, and machine learning concepts. For the first index, HL-modified PLP coefficients and their statistics were used as the feature set, which was subsequently mapped to the predicted quality scores using support vector regression (SVR). For the second index, HL-impacted gammatone auditory filterbank energies and their second-order statistics constituted the feature set, which was again mapped using SVR. Two databases involving HA recordings were collected and utilized for the evaluation of the robustness and generalizability of the two indices. Experimental results showed that the index based on the gammatone filterbank energies not only correlated well with HA quality ratings by hearing impaired listeners, but also exhibited robust performance across different test conditions and was comparable to the full-reference HASQI performance.