Explore the concept of functional specificity in the human brain through this insightful one-hour lecture by MIT's Nancy Kanwisher. Delve into three key questions surrounding brain function, examining scanning techniques, CNN-based models, and image-computable models. Investigate the ideal feature space, shared processing stages, and the potential of larger networks. Analyze behavioral RDMs and discuss the specialization of face recognition, spontaneous segregation, and long-range connectivity. Engage with thought-provoking inquiries about optimal stimuli generation, multiple model comparisons, and the necessity of specific functions in brain processing.
Functional Specificity in the Human Brain - What, Whether, and Why?
Overview
Syllabus
Introduction
Three Questions
Scanning
Cnnbased model
Does it work
Results
Image computable models
Predicted stimuli
What have we done
The Ideal Feature Space
Can Processing Stages Be Shared
What About a Larger Network
Behavioral RDMs
Summary
Questions
Is faces special
Is there a way to test this
Can we probe the network to generate optimal stimuli
Can we have three different models
Do we need specific functions
Face recognition
Spontaneous segregation
Longrange connectivity
Taught by
MITCBMM
