Modelling the Perception of Colour Patterns in Vertebrates with HMAX

• DOI: 10.1101/552307
• Publication Date: 2019-02-19T05:35:11Z
• Authors: Julien P. Renoult; Bastien Guyl; Tamra C. Mendelson; Alice Percher; Jérôme Dorignac; Fredéric Geniet; Molino François
• Keywords: Neural coding; Predictive coding; Computational model

Abstract In order to study colour signals as animals perceive them, visual ecologists usually rely on models of vision that do not consider patterns–the spatial arrangement features within a signal. HMAX describes family are used pattern perception in human research, and which have inspired many artificial intelligence algorithms. this article, we highlight the sensory brain mechanisms modelled widespread, occurring most if all vertebrates, thus offering wide range applications ecology. We begin with short description neural emphasizing similarities processes across species. Then, provide detailed HMAX, highlighting how model is linked biological vision. further present sparse-HMAX, an extension includes sparse coding scheme, make even more biologically realistic tool for estimating efficiency information processing. illustrative analysis, then show performs better than two other reference methods (manually-positioned landmarks SURF algorithm) between faces nonhuman primate This manuscript accompanied MATLAB codes efficient implementation sparse-HMAX can be flexibly parameterized non-human vision, goal encourage adopt tools from computer computational neuroscience.