作者: Arnaud Robert , Jan L. Eriksson
DOI: 10.1121/1.427935
关键词: Hair cell 、 Computer science 、 Filter (signal processing) 、 Cochlea 、 Acoustics 、 Stimulus (physiology) 、 Perceptual Masking 、 Cochlear nerve 、 Auditory perception 、 Noise
摘要: This paper presents a phenomenological model of the cochlea. It consists bank nonlinear time-varying parallel filters and an active distributed feedback. Realistic filter shapes are obtained with all-pole gamma-tone (APGF), which provides both good approximation far more complex wave propagation or cochlear mechanics models very simple implementation. Special care has been taken in modeling properties order to mimic responses cochlea stimuli. As result, reproduces several observed phenomena including compression, two-tone suppression, suppression tones by noise. The feedback, based on physiological evidence from outer hair cell (OHC) functioning, controls damping parameter APGF low-side high-side suppression. Responses stimuli as well study model's parameters also presented. Areas application this type include understanding signal coding auditory nerve, development hearing aids, speech analysis, input neural higher centers.