Vision-to-value transformations in artificial neural networks and human brain
Trung Quang Pham, Takaaki Yoshimoto, Haruki Niwa, Haruka K Takahashi, Ryutaro Uchiyama, Teppei Matsui, Adam Keith Anderson, Norihiro Sadato, Junichi Chikazoe
This study compares CNNs to human brain activity to show that vision-to-value transformation is hierarchical, mirroring neural pathways from primary visual cortex to transmodal regions.
bioRxiv · Mar 18, 2021 · PreprintNeuroscience