A novel binaural speech enhancement algorithm for hearing impaired listeners

With the advent of wireless ear to ear communication, binaural speech enhancement becomes feasible in commercial products such as hearing aids. The first such an example is a fixed binaural beam former that attenuates noise from the nonlook direction while preserving the speech in the look direction. While such an algorithm has shown significant benefits in the lab, it has not produced the wow effect in the field as we have all expected. In this paper, we will discuss some of the practical issues limiting the benefits of such an algorithm in the field. In addition, a novel binaural speech enhancement algorithm is proposed by balancing robustness along with optimal performance. Its benefits will be demonstrated by comparing it with the leading binaural speech enhancement algorithms. Furthermore, future research directions will be discussed.