The abstract is read at the publisher; the summary is JClub's.
Perceived images across five visual illusion classes show higher gray-level Shannon entropy than the corresponding retinal images, indicating added structure beyond the original input.
- Why it matters: Visual illusions challenge the idea that perception faithfully copies retinal signals, implying the brain must infer missing or ambiguous information.
- What they did: The study computed gray-level Shannon entropy for original versus perceived images in five illusion types: edge detection, surface filling-in, brightness perception, size perception, and the Hermann grid.
- The result: In every illusion class, perceived-image entropy exceeded real-image entropy, supporting a unified view that constructive inference adds information for global coherence.
The findingWhy it mattersWhat they didThe result
AI summary of the abstract · may contain errors · not medical advice · read the paper.