UCL Ear Institute

Research Summary

The many sound-generating sources within the environment add up into one combined wave-form
that enters the ear. In order to make sense of the world, a listener has to be able to separate this input into representations of the difference objects in the scene, determine their location in space, recognize them, and react appropriately. This generally occurs automatically and without explicit effort: We follow a conversation in a noisy restaurant, or close our eyes in the park and still perceive the

birds on the tree, the dog barking in the corner and the children playing in the playground. The main objective of my work is to understand the processes by which such a representation is created by the brain and how it is maintained - the processes by whichauditory sensory information is converted into a perceptual representation of our surrounding environment. Although most of the experimental questions that I pose are relevant for all sensory modalities, I choose to address them in the auditory domain both because auditory processing is much less understood than vision, for example, and because two of the sensory experiences that are considered to be most uniquely human – Speech and Music – are primarily auditory in nature.

My methodology is based on a combination of functional brain imaging and psychophysics. By studying how brain responses unfold in time, I explore how representation that are useful for behavior arise from sensory input and dissociate automatic ,stimulus-driven, processes from those that are affected by the perceptual state, task and goals of the listener. Examples of the questions I address in my experiments are: How do listeners detect the appearance or disappearance of new auditory objects (sound sources) in the environment? What makes certain events ‘pop-out’ and grab listeners’ attention even when it is focused elsewhere while the detection of other events requires directed listening? How are listeners able to focus attention on one sound in a mixture? Are listeners able to selectively ignore a moment in time? How does visual input affect how listeners process auditory information?