Visual Nonclassical Receptive Field Effects Emerge from Sparse Coding in a Dynamical System

Abstract

Extensive electrophysiology studies have shown that many V1 simple cells have nonlinear response properties to stimuli within their classical receptive field (CRF) and receive contextual influence from stimuli outside the CRF modulating the cell's response. Models seeking to explain these non-classical receptive field (nCRF) effects in terms of circuit mechanisms, input-output descriptions, or individual visual tasks provide limited insight into the functional significance of these response properties, because they do not connect the full range of nCRF effects to optimal sensory coding strategies. The (population) sparse coding hypothesis conjectures an optimal sensory coding approach where a neural population uses as few active units as possible to represent a stimulus. We demonstrate that a wide variety of nCRF effects are emergent properties of a single sparse coding model implemented in a neurally plausible network structure (requiring no parameter tuning to produce different effects). Specifically, we replicate a wide variety of nCRF electrophysiology experiments (e.g., end-stopping, surround suppression, contrast invariance of orientation tuning, cross-orientation suppression, etc.) on a dynamical system implementing sparse coding, showing that this model produces individual units that reproduce the canonical nCRF effects. Furthermore, when the population diversity of an nCRF effect has also been reported in the literature, we show that this model produces many of the same population characteristics. These results show that the sparse coding hypothesis, when coupled with a biophysically plausible implementation, can provide a unified high-level functional interpretation to many response properties that have generally been viewed through distinct mechanistic or phenomenological models.

Optimal Sparse Approximation With Integrate and Fire Neurons

“…The sparse coding model has been shown to account for many observed response properties in the primary visual cortex (V1), including classic receptive field structure 3,4 and nonclassical response modulations. 5 Furthermore, recent electrophysiology studies also appear to support this coding model. [6][7][8] While there is some debate about the role of sparse coding in biophysical systems, 9,10 some of these criticisms themselves have led to rebuttals in the literature.…”

Response Properties of Single Neurons Predicted by Sparse Representation

“…Recently, some other studies have reported that several classical receptive field (CRF) and non-classical receptive field (nCRF) response properties of single neurons are well simulated by sparse representation [22]- [24]. To further testify this hypothesis, we use a neurally plausible sparse representation model to simulate the emergence of several known simple cell response properties.…”

Constrained inference in sparse coding reproduces contextual effects and predicts laminar neural dynamics

“…However, most of the effects observed in experiments are caused by stimuli extending far beyond the 75 range of the recorded neuron's input fields Polat et al (1998);Walker et al (2000); Mizobe et al (2001). 76 Hence the mechanism put forward by this model Zhu and Rozell (2013) can only be a valid explanation for 77 a small part of these effects, covering situations in which the surround is small and in close proximity to 78 the cRF. This observation raises the important question, how sparse coding models have to be extended 79 to better reflect cortical dynamics and anatomical structure.…”

“…Recently, Zhu and Rozell reproduced a variety of key effects such as surround suppression, cross-67 orientation facilitation, and stimulus contrast-dependent ncRF modulations Zhu and Rozell (2013). In 68 their framework, small localized stimuli are best explained by activating the unit with the optimal match 69 between its input field ('dictionary' vector).…”

“…given the neurobiological constraints on anatomy and neuronal dynamics. In this regard, the neural 392 implementation proposed by Rozell et al (2008) provided a significant advance, since it can explain a range 393 of contextual effects Zhu and Rozell (2013) with a neural population dynamics that requires only synaptic 394 summation and can also be extended to obey Dale's law Zhu and Rozell (2015). But this model still 395 presents a fundamental, conceptual difference to visual cortex: there are no interactions between neurons 396…”