Category selectivity in human visual cortex: Beyond visual object recognition.

Area of Science:

• Neuroscience
• Cognitive Science
• Computational Neuroscience

Background:

• Human ventral temporal cortex exhibits categorical organization, with distinct regions selectively responding to faces, bodies, tools, scenes, and words.
• Traditional explanations link this category selectivity to hierarchical visual object recognition systems and deep neural network models.

Purpose of the Study:

• To challenge the traditional object recognition framework as the sole explanation for category selectivity in visual cortex.
• To propose an alternative framework where category selectivity arises from the demand for specific representations supporting diverse cognitive functions.

Main Methods:

• Review and re-evaluation of existing theories on category selectivity.
• Consideration of category-selective region activation in tasks beyond visual recognition (e.g., navigation, social cognition).
• Examination of connections between category-selective regions and broader domain-specific networks.

Main Results:

• Category-selective regions are activated in non-visual tasks and even without visual experience.
• These regions show close functional connections with broader domain-specific networks.
• The representations within these regions are likely idiosyncratic and not purely visual, tailored for specific tasks.

Conclusions:

• The object recognition framework alone is insufficient to explain category selectivity.
• Category selectivity is better understood as driven by the demand for specific, task-relevant representations.
• This broader perspective necessitates new experimental and computational approaches to study visual cortex organization.