(Leaders: Mountain and Shamma)
A primary aim of the proposed projects is to demonstrate the applicability and advantages of a wide range of auditory-based signal processing algorithms developed within the last few years. Another is to adapt and formulate new algorithms based on new physiological findings and biophysical models of auditory function. In broad terms, sound processing in the auditory system can be subdivided into: (A) the peripheral stages which include the cochlear analysis and transformations of sound into nervous activity in the auditory-nerve; (B) the brainstem and midbrain structures performing the subsequent processing in the cochlear nucleus, superior olive, and up to the inferior colliculus, to extract elementary spectro-temporal and spatial features for various higher level tasks; (C) the central stages which involve the thalamic and cortical centers where multiple spectral and temporal features from convergent binaural and monaural pathways combine to create higher level representations. In each of these three sections, we first describe research into the biophysical models, and then their abstractions as signal processing algorithms. Section (D) briefly describes the physiological experiments to support the biophysical models. Finally, we outline in section (E) how it has been possible in recent years to acquire a deeper appreciation of the unique properties of auditory processing and its strengths in the context of the generalized theory of time-frequency analysis.