Enhancement of Hearing Aid Processing Via Spatial Spectro-Temporal Post-Filtering with a Prototype Eyeglass-Integrated Array

The approach to directional speech enhancement in this paper involves combining adaptive beamforming and real-time capable spectro-temporal post-filtering. Using a prototype listening device with an eyeglass-integrated array, we implement and evaluate our Short-Time Target Cancellation (STTC) processing, which leverages phase differences to compute a spatially-selective (i.e., directional) time-varying spectral gain that can be used either to filter the binaural signals at the ears or as a postfilter for hearing aid processing. In this study, the STTC processing is both compared with, and combined with, an established hearing aid algorithm; the two approaches are compatible and our evaluation results, using instrumental measures and multi-talker mixtures in both anechoic and reverberant settings, indicate that their combination has an additive effect. The approach taken here, using an eyeglass-integrated array to compute a spatially-selective spectro-temporal post-filter, can be used to enhance hearing aid performance in especially challenging acoustic environments.