Abstract:This modeling study proposes a novel theory for how V4 neurons embody selectivity to the varying degrees of curvature in diverse receptive field arrangements reported in previous recording studies. Cortex is modeled as having neurons with simple, fixed arrangements of connections to topologically organized afferent inputs repeating in a rotational tiling: V1 aggregates points into selectivity for lines and edges at a variety of rotations; V2 aggregates orientated segments into selectivity for corners at a variety of angles and rotations; V4 aggregates corners into curvature selectivity. It is demonstrated that a simple, systematic approach to curvature selectivity can explain a wide variety of past observations. The model is implemented in approximately 900,000 integrate and fire units obeying Dale's law in standard cortical micro-circuits. The simple patterns of afferent connections repeat in topologically organized layers. and a novel method for simulating large numbers of integrate and fire units with tensor programming and GPU hardware is employed. By combining convolution with the time course of post-synaptic potentials (PSP), computation occurs in a feedforward cascade of synchronized spikes. This leaves open the question of connection formation as the model has relatively few parameters which are tuned with the Covariance Matrix Adaptation Evolution Strategy (CMA-ES) plus manual fine tuning. This is sufficient to qualitatively reproduce a wide variety of V4 recordings and demonstrates that a single mechanism of curve selectivity can give rise to the diverse repertoire of behaviors observed in V4. A synthesis of hierarchical and convolutional network models, this study adds novel elements to both: In comparison to previous hierarchical models there is a novel V4 mechanism and a computation based on single spikes using tensor convolution as the computational engine. In comparison to previous convolution models there is a much higher degree of biological realism and a novel unsupervised approach to connection formation.

A biologically inspired computational model of human ventral temporal cortex

Principles Governing the Topological Organization of Object Selectivities in Ventral Temporal Cortex

Computational Constraints Underlying the Emergence of Functional Domains in the Topological Map of Macaque V4

Distributed representation of objects in the human ventral visual pathway

Organic Convolution in The Ventral Visual Pathway Can Explain the Variety of Shape Tuning in Area V4

Modeling Category-Selective Cortical Regions with Topographic Variational Autoencoders

A unifying framework for functional organization in early and higher ventral visual cortex

A Map of Object Space in Primate Inferotemporal Cortex

An object-topic map in primate inferotemporal cortex

Topological Receptive Field Model for Human Retinotopic Mapping

Vicarious Somatotopic Maps Tile Visual Cortex

Mechanisms underlying category learning in the human ventral occipito-temporal cortex

A multi-scale cortical wiring space links cellular architecture and functional dynamics in the human brain

Mesoscopic landscape of cortical functions revealed by through-skull wide-field optical imaging in marmoset monkeys

Functional connectomics reveals general wiring rule in mouse visual cortex

Mesoscale organization of ventral and dorsal visual pathways in macaque monkey revealed by 7T fMRI

Brain connectomes come of age

Functionally Specific and Sparse Domain-Based Micro-Networks in Monkey V1 and V2

Balancing Selectivity and Generality in Object Recognition through Structured Interconnectivity.

Connectome-based models of feature selectivity in a cortical circuit

Deep learning-driven characterization of single cell tuning in primate visual area V4 unveils topological organization