Visual number sense for real-world scenes shared by deep neural networks and humans

Published: 17 April 2023| Version 1 | DOI: 10.17632/8j7fwfk8g3.1
, Lin Chen,
, Liu Zuxiang


This study investigated the number perception of scenes using AlexNet. The study findings show that numerosity can be represented by the pattern of group activation of the category layer units. We believe that our study makes a significant contribution to the literature because it is the first to show that a distinct feature in visual environments, revealed by deep neural networks, can modulate human numerosity perception, supported by a group-coding mechanism.



Visual Perception, Deep Neural Network