Medical Physics

Technical and Perceptual Evaluation

In February 2013 the World Health Organization (WHO) states that 360 million people worldwide (5% of the world's population) have to deal with disabling hearing loss. For adults a disabling hearing loss refers to a hearing loss greater than 40 dB in the better hearing ear.

To communicate in acoustically difficult situations, with for example more than one speaker (the cocktail party problem), is a difficult task for every person. To challenge this task the healthy human auditory system is able to use several signal properties, monaural cues as for example spectral profiles or level modulations but also binaural cues as level or phase differences between the left and right ear. Hearing impaired listeners are significantly limited in using these important cues for sound source localization and auditory scene analysis.

In modern hearing aids advanced digital signal processing is used to compensate for these limitations. The main goal is to increase the speech intelligibility while not degrading the signal quality for each patient. In the process of reaching this goal the complexity (number of free parameters) of the developed algorithms is increasing continually, which makes the effective optimization and individualization an important challenge.

The main aim of the TaPE group is to help by broadly characterizing, both individually and together, modern signal processing strategies and the user of those technologies. Therefore we utilize physical measurements, predictions of effective psychoacoustic models and subjective psychoacoustic. The combination of these three different approaches allows us to describe not only how the algorithms are actually manipulating the signal but also to evaluate the effectiveness in the model as well as the efficiency in the real world.

Our work is supported by different public sector funding bodies, mainly the Bundesministerium für Bildung und Forschung (BMBF) and the Deutsche Forschungsgemeinschaft (DFG).

Keywords

psychoacoustics, auditory perception, speech intelligibility, monaural, binaural, loudness, pitch, localization, amplitude modulation, fine structure, sound, hearing impairment, model,  hearing aids, algorithms