Dimension-based Attention in Learning and Understanding Spoken Language

AbstractThe acoustic world is variable and messy. Signals that allow listeners to identify and/or disambiguate important auditory objects (an unusual rustle of leaves in a darkening forest, your winning bingo number) will likely share many acoustic characteristics with other sounds in the environment. What is more, the most diagnostic auditory dimensions for detecting the ominously rustling leaves or categorizing a crucial syllable may be obscured, distorted, or missing. Just as often, the diagnostic acoustic dimensions may change completely across contexts.

Return to previous page