Mid-level visual features engage V4 to support sparse object recognition under uncertainty
Leyla Loued‐Khenissi, Elsa Scialom, Antoine Lutti, Michael Herzog, Bogdan Draganski
This study shows that mid-level visual features, specifically curve segments, enable better object recognition under uncertainty by engaging V4, offering advantages over phosphene-like dots in cortical prostheses.
bioRxiv · Oct 28, 2025 · PreprintNeuroscience