Two cortical routes link scene context and animate content

What the study found: The study identified two distinct processing routes during scene perception in the human cortex: a ventromedial pathway specialized for scene layout and environmental context, and a lateral occipitotemporal pathway selective for animate content.

Why the authors say this matters: The authors conclude that these findings refine classical visual-stream models by showing a distributed cortical network with separable representational routes for context and animate content during scene perception.

What the researchers tested: The researchers applied representational similarity analysis to 7T fMRI data collected during natural scene viewing. They compared representational geometry shared across individuals with hierarchical features from vision and language neural networks across model layers, and integrated these comparisons with representational connectivity between cortical regions.

What worked and what didn't: Vision models aligned with shared structure in both routes. Language models corresponded mainly with the lateral pathway and showed negative alignment in early visual and ventral cortex.

What to keep in mind: The available abstract does not describe specific limitations, caveats, or scope constraints beyond the study's focus on natural scene viewing and the analyses used.

Key points

• Two cortical routes were identified during scene perception.
• One route was linked to scene layout and environmental context.
• The other route was selective for animate content.
• Vision models aligned with shared structure in both routes.
• Language models aligned mainly with the lateral pathway and negatively with early visual and ventral cortex.