What question did this study set out to answer?

This research investigates how the organization of local object fragments in the pigeon brain affects visual perception and accuracy.

April 19, 2026动物学研究Open Access

Prototype-based geometry of local object patches in the pigeon Mesopallium Ventrolaterale

View Full Paper

Ask AI

Bookmark

Authors

Prototype-based learning reveals how object fragments affect accuracy in pigeon vision, suggesting new brain-inspired models.

This research investigates how the organization of local object fragments in the pigeon brain affects visual perception and accuracy.
Combined whole-part image categorization with population electrophysiological recordings
Used CEBRA-time to embed neural time series into low-dimensional space
Defined neural prototypes from population responses to whole images
Projected local patches into neural space to assess distances to prototypes
Introduced local prototype learning mechanisms for category matching based on proximity.
Animal categories formed compact clusters around neural prototypes in MVL
Greater prototype distance associated with lower behavioral accuracy
Patches closer to prototypes had more diagnostic semantic parts (e.g., heads, faces)
Local prototype learning reproduced clustering and behavior effects seen in MVL.

Xiaoteng et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark

April 19, 2026动物学研究Open Access

View Full Paper

Ask AI

Bookmark

Authors

Prototype-based learning reveals how object fragments affect accuracy in pigeon vision, suggesting new brain-inspired models.

This research investigates how the organization of local object fragments in the pigeon brain affects visual perception and accuracy.
Combined whole-part image categorization with population electrophysiological recordings
Used CEBRA-time to embed neural time series into low-dimensional space
Defined neural prototypes from population responses to whole images
Projected local patches into neural space to assess distances to prototypes
Introduced local prototype learning mechanisms for category matching based on proximity.
Animal categories formed compact clusters around neural prototypes in MVL
Greater prototype distance associated with lower behavioral accuracy
Patches closer to prototypes had more diagnostic semantic parts (e.g., heads, faces)
Local prototype learning reproduced clustering and behavior effects seen in MVL.

Xiaoteng et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark