Object category structure in response patterns of neuronal population in monkey inferior temporal cortex

Area of Science:

• Neuroscience
• Cognitive Science
• Computational Neuroscience

Background:

• Human object categorization is hierarchical and essential for daily function.
• The inferior temporal (IT) cortex is a key brain region for object recognition.

Purpose of the Study:

• To investigate how object categories are represented by neural activity in the monkey IT cortex.
• To explore the hierarchical structure of object representations in the IT cortex.

Main Methods:

• Recorded activity from over 600 neurons in the IT cortex of monkeys.
• Presented over 1,000 natural and artificial object images during a passive fixation task.
• Analyzed population activity patterns to identify categorical representations.

Main Results:

• Population activity in the IT cortex reflects the hierarchical structure of object categories.
• Distinguishable clusters emerged for animate versus inanimate objects.
• Further subdivisions within animate objects (e.g., bodies, faces) were represented.
• Hierarchical clustering revealed specific groupings, such as primate faces and different animal body types.

Conclusions:

• The IT cortex exhibits an inherently categorical neural code.
• Population responses in the IT cortex reconstruct intuitive object category structures.
• Category representation is distributed across neurons, including those with strong and weak responses.