We explore the interaction between local-global informa-
tion processing in visual perception, leveraging a visual
phenomenon known as crowding, whereby the perception
of a target stimulus is impaired by the presence of nearby
ankers. The majority of established models explain
the crowding e?ect in terms of local interactions. How-
ever, recent experimental results indicate that a classical
crowding e?ect, the deterioration in the discrimination
of a vernier stimulus embedded in a square, is alleviated
by the presence of additional
anker squares (\uncrowd-
ing"). Here, we propose that crowding and uncrowding
arise from cortical inferences about hierarchically orga-
nized groups, and formalize this concept using a hierarchi-
cal Bayesian model. We show that the model reproduces
both crowding and uncrowding for
anked vernier discrim-
ination. More generally, the model provides a normative
explanation of how visual information might simultane-
ously
ow bottom-up, top-down, and laterally, to allow
the visual system to interactively process local and global
features in the visual scene.