Dynamical Model of Auditory Scene Analysis

F0 estimation by GrFNN model (CogMIR 2016)

Auditory scene analysis refers to the organization of sound elements mixed in an acoustic signal into separate perceptual objects. We explain auditory scene analysis as dynamic pattern formation in nonlinear oscillatory systems. Our current focus is the segregation of concurrent harmonic sounds such as musical tones and speech sounds. The emergent pattern of mode-locked synchronization between neural oscillators provides a biologically realistic account of the grouping and segregation of harmonics of multiple concurrent F0s.

Kim, J. C., & Large, E. W. (2016). Multiple F0 estimation by gradient frequency neural networks. The 6th Annual Seminar on Cognitively Based Music Informatics Research, New York, NY. [slides]

Kim, J. C., & Large, E. W. (2016). A nonlinear dynamical systems approach to auditory scene analysis. The 14th International Conference for Music Perception and Cognition, San Francisco, CA. [poster]