A microphone for personal communication is placed inside the external auditory canal of a user. A pair of ear-muffs equippedwith Active Noise Control (ANC) are fitted onto the head of the user. The combination of passive and active noise reduction makes communication more reliable when performed in noisy environments. In this paper, a Hidden Markov Model (HMM) speech recognition system is used to assess the quality of speech signals recorded using an ear-mic combined with ANC. In addition, speech recognition itself can be of interest in many situations. For example, many functions in a vehicle can be controlled by instructions spoken by the user and speech controlled functionality improves safety. Hence, speech recognition is a growing field in vehicular technology. If the speech intelligibility and by that the speech recognition robustness could be increased in a speech recognition controlled system implemented in a vehicle, the safety would be further improved.

En taligenkännare används för att påvisa en förbättring av talkvalitet då en mikrofon för kommunikationsändamål placeras i hörselgången på en användare och detta kombineras med ett par aktiva hörselkåpor.