RWTH Aachen
University
Institute for Communication
Systems and Data Processing
Skip to content
Direkt zur Navigation
Home
Home

Publications – Details

Opening the Ear Digitally – Natural perception of the own voice while using hearing devices

Authors:
Stefan Liebich, Peter Jax, and Peter Vary
Book Title:
45th Erlanger Kolloquium
Venue:
Erlangen
Event Date:
22.02-23.02.2018
Organization:
Sivantos
Date:
February 2018
Note:
Invited talk
Language:
English

Abstract

The perception of the own voice is severely altered when occluding the ear canal with a hearing device, such as communication headsets or hearing aids. This is called the occlusion effect, which is a usual complaint by the users. It manifests as an amplification of low frequency components and an attenuation of high frequency components of the own voice. The occlusion effect can be reduced by passive means with deep insertion of the hearing device into the ear canal, which reduces comfort, or ventilation holes, which increase the problem of feedback between inner loudspeaker and outer microphone. Especially for large sound amplification gains, e.g. for large hearing losses, this imposes a huge problem. An alternative approach is to attenuate the over-amplified low frequencies by Active Noise Cancellation (ANC), using acoustical counter waves, and to increase high frequencies by Active Signal Injection (ASI). The hearing devices need an inner microphone facing the eardrum and an outer microphone recording the ambient sound. The inner microphone is used for the ANC to create the counter signal and the outer microphone is used for the ASI to create an equalized ambient sound signal. Both signals are played together via the loudspeaker/receiver. In this talk, we are presenting an Active Occlusion Cancellation for hearing devices. The algorithms are implemented and evaluated in a real-time system for the application of communication headsets. A fixed feedback controller designed with H-infinity optimization performs the ANC. It is further adjusted by an adaptive approach to ensure robustness and performance in the presence of time-varying acoustic conditions. Furthermore, the ambient signal induced by the ASI, to enhance high frequency components and create a natural perception of the own voice. The developed methods lay the foundation for the application in the field of hearing aids, with specific challenges due to the acoustic receiver, hearing aid fitting and energy consumption. The challenges will briefly be outlined in the talk.