- These “grandmother neurons” do not activate when shown individual facial features of the person
- You would need enough neurons for each person and object you can see, which might be more neurons than we have

University of Michigan - Ann Arbor

This theory explains that visual input is processed hierarchically.
- Theoretically, lower-level neurons would process different facial features and send them to higher-level neurons, and you would end up with "grandmother neurons," "Halle Berry neurons," etc.
    - Have individual neurons that are responsible for recognizing each visual object

Hierarchical model of vision processing

Breedlove, S. & Watson, N. (2017). Behavioral neuroscience, eighth edition. Oxford University Press.

Learn Before

Related