Evaluation of noise reduction in digital hearing aids in situations with multiple signal sources
Marlitt Frenz, Alfred Mertins, Hendrik Husstedt · Proceedings of the International Symposium on Auditory and Audiological Research · 2017
Features of modern hearing aids such as the digital noise reduction and directional microphones can enhance speech. One way to evaluate the effect of these features is to measure the increase of the signal-to-noise ratio (SNR). To this end, the method of Hagerman and Olofsson is often used. However, only two signals can be distinguished with this method, e.g., speech and noise. Since many realistic situations include more than two signals, an extension of the method of Hagerman and Olofsson for an arbitrary number of signals is introduced. To proof the concept, this extended method is applied to a setup with 9 different signals presented by 8 speakers. This study considers a separation of speech and noise for 8 signal sources. All speakers are positioned around a hearing aid on a circle with a radial distance of r = 1m and an angular distance of 45◦ between 0◦ and 360◦. Speech is presented from 0◦ and noise from all 8 directions (0◦,45◦, ...,360◦). With this setup, a stateof- the-art hearing aid is analysed for different settings where the digital noise reduction and/or the directional microphones are turned on or off. As a result, the SNR for all directions can be investigated individually. This demonstrates the practicability of the extended method.