Analysis and development of speech enhancement features in cochlear implants

Orateur : Nicolas Furnon

Date : le 18 octobre 2018 à 10h30 – C005

Résumé :

Cochlear implants (CIs) are complex systems developed to restore the hearing sense to people with profound auditory loss. These solutions are efficient in quiet environments but adverse situations remain very challenging for CI users. A range of algorithms are implemented in the processors by the different manufacturers to improve speech intelligibility (SI) of the CI users. Among them, beamformers and automatic gain controls (AGCs). We will present the effects of these algorithms on the SI and develop them in order to increase their performances. The studies revealed that the beamformers are efficient solutions which can improve the signal-to-noise ratio (SNR) up to 9 dB in noisy environments. As for the AGC, a system adapting the output level of the signal, it was thought of using a multichannel system instead of the single-channel system that is currently used. The new resulting algorithm proved to be an efficient noise-segregating feature. The simulations carried out to evaluate it were confirmed in listening tests. The speech reception threshold (SRT), which is the SNR required for 50% understanding of the target material, decreased by 9.8 dB.