Signal processing and signal enhancement of speech and audio signals are among the main strengths of the Oldenburg Branch for Hearing, Speech and Audio Technology HSA. Tasks include the optimization of disrupted signals as well as the development of recording and processing technologies. This allows audiological data to be used for acoustic sensors even in a challenging environment, such as intense ambient noise. Apart from the enhancement of audio and speech signals the localization of sound sources as well as automatic speaker recognition are key research priorities for the Fraunhofer IDMT in Oldenburg.
In security applications, relevant sounds can be separated from less relevant ones by reducing background noise in preparation for acoustic monitoring. Specific examples are the recognition of breaking glass or the use of spray cans for combating break-ins and vandalism, but also the detection of drones. The reduction of background noise is a fundamental element of solutions in the area of speech recognition and voice control, for example in smart home applications or conference systems.
In addition, the design of personal hearing systems is enhanced through the individual reduction of the level of reverberation. Automatic volume regulation and the personalized adjustment of high and low signals ar other field ofs optimization. Both manufacturers of audio system equipment, multimedia devices or hearables as well as users with normal or impaired hearing profit. The localization of sound sources is a technology also used, for example, in the automotive sector for recognizing the direction of approaching emergency vehicles.