Psychoacoustical modelling of auditory object perception in humans
Similar to the visual system which groups the environment into different visual objects such as e.g., a chair or a table, the auditory system distinguishes different auditory objects in a complex acoustical environment. In real acoustical environments, an auditory object usually corresponds to the sound of a particular sound source and it is generally assumed that characteristics of these natural sound sources are used as object binding cues. Motivated by results of the previous funding period, a realistic nonlinear model of the cochlea will be used to quantify its contribution to the processing of coherent envelope fluctuations across frequencies which is a common property of natural sound sources. In a second step, a physiologically motivated model will be developed that is sensitive to several object-binding cues. The combination of binaural and monaural cues will be investigated in free field and under headphone conditions. This is especially interesting since the object binding in these conditions work on different time scales: The typical interaural time differences providing information about the spatial location of a source are shorter than one millisecond whereas the time scale for coherent envelope fluctuations across frequency is at least ten times longer. Since there is increasing evidence that the auditory system uses dynamic changes as additional object-binding cues, experiments will be performed investigating the sensitivity to dynamic changes in spectro-temporal patterns as well as dynamic variations of the binaural cues (e.g. a moving sound source). The results will provide insights into the combination of different object-binding cues in real acoustical environments. The comparison of the results and the predictions of the bottom-up model provide insights into the relative contribution of bottom-up and top-down processes and will be used for an extension of the model including also top-down processes.
auditory objects, psychoacoustical modelling
Die Daten werden geladen ...
Keine Ergebnisse gefunden, bitte ändern Sie Ihre Suchanfrage.