We revealed the algorithms and mechanisms of rapid luminance gain control in Drosophila, which is a critical computation that results in stable visual processing in highly dynamic light conditions. We identify the full circuitry that achieves stable visual processing, and use a combination of experiments and theory to show that narrow spatial pooling luminance signal achieves luminance gain and is tuned to the statistics of natural visual scenes.
While we studied the implementation of rapid luminance gain control in the fly, all visually guided animals and machines are confronted with the same (rapid luminance change) challenges, arguing that similar mechanisms will also aid (or could aid) stable vision across systems.
Read the full article published in Nature Communications here
