In the APP lab, we study auditory perception broadly with special emphases on perceptual organization, perceptual learning, and speech perception. Our research explores the influences of predictable versus unpredictable signal properties on perceptual behaviors such as discrimination, identification, and recognition. We explore these effects of predictability through multiple avenues: how reliable spectral features in an acoustic context affect identification of a subsequent target sound; how redundancy between acoustic properties in novel complex sounds (and the degree to which that redundancy is violated) shapes their discriminability; and, the perceptual importance of information-bearing acoustic change for speech recognition in normal hearing and simulations of cochlear-implant processing. We utilize a wide array of psychoacoustic tasks complemented by computational models of auditory perception.
Our research has a wide range of influences: auditory neuroscience and physiology, as we explore perceptual consequences of neural response properties; neural network modeling and computational perception, to understand principles of perceptual organization; high-level perception in other modalities (particularly vision), to understand fundamental tenets of perception most broadly; and, information theory and the efficient coding hypothesis, which elegantly characterize a great deal of perception through a simple, immutable tenet: the world is filled with structure, and perception excels at extracting and exploiting this structure.
Major themes of our research include:
-how perception detects and exploits statistical regularities in the acoustic environment
-weighting and integrating different sources of acoustic information
-understanding effects of short-term and long-term experience on perception
-common principles in neural and perceptual organization
-implications of efficient neural coding for perception
-computational principles that explain perceptual performance
-characterizing the statistical structure of natural sounds and how this reflects response properties of the auditory system
-acoustic information for speech perception available to listeners with normal hearing as well as individuals who use cochlear implants