Celine Spriet^1,2 (), Emilie Serraille^1,2, Liuba Papeo^1,2, Jean-Rémy Hochmann^1,2; ¹Institut des Sciences Cognitives Marc Jeannerod, CNRS, UMR5229, ²University Lyon1

A remarkable acceleration of visuo-attentional processes has been described in the first months of life up until adulthood. In parallel, throughout the first years of life, infants seem to integrate more and more visual features together leading to the ability to represent more and finer-grained visual categories. Here, we show that these two processes are related: as infants grow older, visual processing gets faster, facilitating the integration of visual features to form categorical object representations. Using frequency-tagging electroencephalography (EEG), we targeted a response associated with the perceptual distinction between animate and inanimate objects in adults, in 4- and 9-month-old infants. Images (640 animate and inanimate objects) from one category (animate) were presented at a regular, base frequency (Fb), interleaved with images from the other category (inanimate) presented at the regular target frequency Ft=Fb/5. Visual categorization was tested at increasingly faster stimulation frequency (for adults (n=36): Fb=6, 12 or 30 Hz; for 4-month-olds (n=64): Fb=4 or 6 Hz; for 9-month-olds (n=64): Fb=6 or 12 Hz). Results revealed that the baseline-corrected response amplitude at Ft (and harmonics), used as a measure of categorization, decreased as the stimulation frequency increased. In 4-month-olds, the categorization response was only observed at 4 Hz; while it was found with a stimulation frequency as fast as 12 Hz in 9-month-olds, and 30 Hz in adults. The 4-month-olds’ response at 4 Hz was equivalent to the 9-month-olds’ response at 12 Hz, suggesting that 9-month-olds process categories ~3 times faster than 4-month-olds. These results demonstrate a dramatic acceleration of visual categorization, a process relying on feature integration, in the first year of life, which continues into adulthood. We will propose a model where, as the processing speed accelerates, more and more visual features activated by visual stimulation can be integrated, yielding more efficient and richer categorization.

Acknowledgements: This research was funded by the European Research Council Starting Grant to Liuba Papeo (Grant number: THEMPO-758473), the Labex Cortex grant to Jean-Rémy Hochmann and the Fondation pour la Recherche Médicale - Fin De Thèse Grant (FDT202304016547) to Céline Spriet.