The abstract is read at the publisher; the summary is JClub's.
Object-bound features in natural images produce EEG encoding models with a consistently delayed temporal profile relative to scene-bound features in the human brain.
- Why it matters: Standard fMRI localizers use artificial stimulus separations and cannot capture how object versus scene information unfolds over time in real-world vision.
- What they did: The study built image-grounded, deep neural network-based encoding models to predict human EEG responses from per-image object-bound versus scene-bound features extracted from a large set of high-resolution natural images.
- The result: Distinct temporal profiles were linked to object versus scene typicality of individual image elements, and robustness tests showed temporal separation requires a consistent object–scene feature separation.
The findingWhy it mattersWhat they didThe result
AI summary of the abstract · may contain errors · not medical advice · read the paper.