Authors
Rostami, F., Spriet, C., Op de Beeck, H., Hochmann, J.-R., Papeo, L.
Abstract
Human recognition of objects as animate or inanimate is fast and accurate. However, this task may be challenging for objects with animal-like properties (e.g., presence of eyes/face), albeit being inanimate (e.g., a cow-mug). Lookalike objects provide an opportunity to examine how visual perception resolves categorical ambiguities and whether it exhibits an intrinsic bias to see animacy. During electroencephalography (EEG), we presented healthy humans with images of objects at a regular, rapid frequency of 6 Hz, with every fifth image being an exemplar from the oddball category (i.e., one oddball after four standard images), yielding an oddball frequency of 1.2 Hz. In some conditions, oddball-stimuli were replaced by lookalikes. Periodic visual stimulation should give rise to a distinctive EEG response at 6 Hz. Moreover, if the oddball category elicits a different neural response than the standard category, a distinct response should be observed at 1.2 Hz. This categorization response was found for animate-oddballs among inanimate objects, and vice versa. It was also found for lookalike-oddballs presented among animate or inanimate objects, but it was significantly higher in the first case, implying that lookalike objects were perceived as more similar to inanimate than animate objects. The degree of animal resemblance modulated the amplitude of neural response to lookalikes, without however changing the category boundary. These results - also replicated in an artificial model of human vision - demonstrate that animal resemblance of objects is registered in visual perception, but it does not alter the critical ability to distinguish between what is truly alive and what is not.
Preprint server:
bioRxiv
The authors list and abstract were imported from bioRxiv on 20 Jun 2026.
Advertisement
Stats
- Recommendations n/a n/a positive of 0 vote(s)
- Views 0
- Comments 0