Sound Source Localization for Hearing Aid Applications Using Wireless Microphones
Mojtaba Farmani, Michael Syskind Pedersen, Jesper Jensen · 2018
State-of-the-art hearing aids (HAs) can connect to a wireless microphone worn by a talker of interest. This ability allows HAs to have access to almost noise-free sound signals of the target talker. In this paper, we aim to estimate the direction of arrival (DoA) of the target signal, given access to the noise-free target signal. Knowing the DoA of the target signal enables HAs to spatialize the wirelessly received target signals. The proposed estimator is based on a maximum likelihood (ML) approach and a database of DoA-dependent relative transfer functions (RTFs), and it supports both monaural and binaural microphone array configurations. For binaural configurations, we propose an information fusion strategy, which decreases the number of parameters required to be wirelessly transferred between the HAs. Further, we show that the proposed method not only has lower computational complexity, but also performs better than recent DoA estimators, which have access to the noise-free target signal.