Toward a scalable, binaural hearing device: Acoustics, interaural processing, and scene decomposition
Birger Kollmeier, Marko Hiipakka, Giso Grimm, Tobias Neher, Tobias de Taillez, Jens Schröder, Jörn Anemüller, Volker Hohmann · The Journal of the Acoustical Society of America · 2014
One of the aims of the focused research group “Individualized hearing acoustics” are elements of an assistive listening device that both fits the requirements of near-to-normal listeners (i.e., providing benefit in noisy situations or other daily life acoustical challenges) and, in addition, can be scaled up to a complete hearing aid for a more substantial hearing loss. A review is given on the current work on acoustically “transparent” open-fitting earpieces, algorithms for interaural cue enhancement, and for automatically detecting relevant changes in the acoustical scene that call for an altered processing strategy. The current prototype runs on the binaural, cable-connected master hearing aid (MHA) that includes earpieces that allow for approaching acoustic transparency. This can be achieved by individual calibration of the earpieces using in-situ in-ear measurements and electro-acoustic models of the earpieces and the ear canal. A binaural high-fidelity enhancement algorithm motivated by interaural magnification is evaluated in its benefit for normal and near-to-normal listeners. Finally, several acoustic event detection schemes are compared with respect to their applicability in real-life situations. They decompose the ambient acoustic signal into spectro-temporal basis functions and utilize machine learning classifiers to detect typical natural and artificial events. [Work supported by DFG.]