Statistical inference in the auditory system

NSF Award Search · 01002526DB NSF RESEARCH & RELATED ACTIVIT · $450,028 · view on nsf.gov ↗

Abstract

Every day, our brains seamlessly pick out meaningful information from the cacophony of sounds that enters our ears. This ability to make sense of our acoustic surroundings relies on the brain being able to recognize patterns and predict what might come next. However, in real-world environments, sounds are rarely perfectly predictable, raising the question: how does the brain track information when patterns are uncertain or incomplete? This research seeks to understand how the brain extracts and organizes auditory information from complex, dynamic soundscapes. The work has broad implications for real-world applications, from improved audio technologies and assistive listening devices to shedding light on auditory processing challenges in specific clinical populations. The project explores how the brain infers statistical structure from sound sequences, focusing on the way it builds an internal model of the world to interpret and predict auditory experiences. While past research has shown that the brain can recognize predictable patterns in sound, real-world listening often involves incomplete or stochastic sounds, making it unclear how this predictive process functions in natural environments. The investigator uses a combination of computational modeling, behavioral testing, and neurophysiological (EEG) experiments to study how the brain tracks complex statistical relationships and deeper temporal and multi-feature dependencies in sound sequences to guide perception. Indiv

Key facts

NSF award ID
2444353
Awardee
Johns Hopkins University (MD)
SAM.gov UEI
FTMTDMBR29C7
PI
Mounya Elhilali
Primary program
01002526DB NSF RESEARCH & RELATED ACTIVIT
All programs
Perception, Action and Cognition
Estimated total
$450,028
Funds obligated
$450,028
Transaction type
Standard Grant
Period
07/01/2025 → 06/30/2028