The Camouflage Machine: Optimising protective colouration using deep learning with genetic algorithms

Evolutionary biologists frequently wish to measure the fitness of
alternative phenotypes using behavioural experiments. However, many
phenotypes are complex. For example colouration: camouflage aims to make
detection harder, while conspicuous signals (e.g. for warning or mate
attraction) require the opposite. Identifying the hardest and easiest to
find patterns is essential for understanding the evolutionary forces that
shape protective colouration, but the parameter space of potential
patterns (coloured visual textures) is vast, limiting previous empirical
studies to a narrow range of phenotypes. Here we demonstrate how deep
learning combined with genetic algorithms can be used to augment
behavioural experiments, identifying both the best camouflage and the most
conspicuous signal(s) from an arbitrarily vast array of patterns. To show
the generality of our approach, we do so for both trichromatic (e.g.
human) and dichromat (e.g. typical mammalian) visual systems, in two
different habitats. The patterns identified were validated using human
participants; those identified as the best for camouflage were
significantly harder to find than a tried-and-tested military design,
while those identified as most conspicuous were significantly easier than
other patterns. More generally, our method, dubbed the ‘Camouflage
Machine’, will be a useful tool for identifying the optimal phenotype in
high dimensional state-spaces.