Retinal adaptive mechanisms confer selectivity to homogeneous objects in natural scenes

Adaptive mechanisms play a vital role in retinal sensory neurons, allowing them to transmit information in various contexts. Traditionally, their function has been considered as normalizing neuronal responses to input statistics, such as mean and variance. However, these mechanisms have primarily been studied using artificial stimuli, with their performance on natural stimuli remaining uncertain.

In this study, researchers demonstrated that during ganglion cells' response to natural scenes, adaptive mechanisms alter the feature selectivity of these retinal output cells. To replicate these findings, an artificial neural network model with a bio-inspired adaptive mechanism was employed, which was crucial in predicting cell responses to new sequences of natural images.

This adaptive mechanism specifically tunes certain cell types to react selectively to homogeneous regions nestled within complex visual backgrounds, a capability ideally suited for detecting potential threats. Importantly, adaptive mechanisms do not merely normalize responses but actively facilitate new feature selectivity within the early visual system.