Supplemental Materials Menu
Andrew Francl and Josh McDermott train a deep neural network enhanced with human ears, to understand how sounds are localized in a complex acoustic environment.
Overview
This unit explores models of the neural mechanisms underlying three kinds of intelligent processes: (1) auditory processing, including the spatial localization of sounds and recognition of speech and music; (2) the representation of episodic and spatial memory in the hippocampus; and (3) conscious experience of objects and events in the world.
Videos
Josh McDermott: Understanding auditory cortical computation
Josh McDermott first introduces the main components of the auditory system and early models of auditory processing in the brain. He then asks three fundamental questions: can we build better models of auditory perception and its underlying neural mechanisms, what can we learn from new models about auditory behavior and human auditory cortex, and can new models be exploited to help people hear better?
Matthew Wilson: Hippocampal mechanisms of memory and cognition
The hippocampus is involved in episodic memory, which is the linkage of events, as well as the spatial memory used in navigation, which is the linkage of spatial locations. Both may depend critically on temporal sequence encoding. Matthew Wilson describes key research on the role of the hippocampus in these two kinds of memory, and the neural mechanisms of sequence encoding.
Christof Koch: The sciences of consciousness: Progress & problems
Christof Koch reflects on the history of research on what is consciousness, what is currently known about the nature of conscious experience, and what are possible neural correlates of consciousness. He then introduces the Integrated Information Theory of Consciousness that can explain a range of clinical and laboratory findings, and makes key predictions for future research on consciousness.
Further Study
Online Resources
Additional information about the speakers’ research and publications can be found at these websites:
- Josh McDermott (MIT)
- Matthew Wilson (MIT)
- Christof Koch (Allen Institute)
Readings
Francl, A., McDermott, J. H. (2020) Deep neural network models of sound localization reveal how perception is adapted to real-world environments, bioRxiv, pp. 2020.07.21.214486
Jones M. W., Wilson, M. A. (2005) Theta rhythms coordinate hippocampal-prefrontal interactions in a spatial memory task, PLoS Biology, 3, 2187-2199
Kell, A., Yamins, D., Shook, E., Norman-Haignere, S., McDermott, J. (2018) A task-optimized neural network replicates human auditory behavior, predicts brain responses, and reveals a cortical processing hierarchy, Neuron, 98, 630-644
Koch, C. (2019) The feeling of life itself: Why consciousness is widespread but can’t be computed, The MIT Press, Cambridge
Koch, C. (2004) The quest for consciousness: A neurobiological approach, Roberts and Company Publishers, Englewood, CO
Penagos, H., Varela, C., Wilson, M. A. (2017) Oscillations, neural computations and learning during wake and sleep, Current Opinion in Neurobiology, 44, 193-201
Saddler, M. R., Gonzalez, R., McDermott, J. H. (2020) Deep neural network models reveal interplay of peripheral coding and stimulus statistics in pitch perception, bioRxiv, pp. 2020.11.19.389999
Sanders, H., Ji, D., Sasaki, T., Leutgeb, J. K., Wilson, M. A., Lisman, J. E. (2019) Temporal coding and rate remapping: Representation of nonspatial information in the hippocampus Hippocampus, 29, 111-127
Tononi, G., Koch, C. (2015) Consciousness: Here, there and everywhere, Philosophical Transactions of the Royal Society B, 370: 20140167