Abstract
Evolutionary biologists frequently wish to measure the fitness of alternative phenotypes using behavioral experiments. However, many phenotypes are complex. One example is coloration: 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 coloration, but the parameter space of potential patterns (colored 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 behavioral 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 dichromatic (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 to find 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.
Highlights
Evolutionary biologists frequently wish to measure the fitness of alternative phenotypes using behavioral experiments
General Linear Mixed Models (GLMMs) showed that trials became significantly harder over time when optimizing for concealment, while optimizing for visibility yielded easier to find targets (Table 1)
A positive estimate coupled with a significant P-value suggested that targets became harder to see over the course of the experiment, while negative estimates indicated that targets became easier to see
Summary
Evolutionary biologists frequently wish to measure the fitness of alternative phenotypes using behavioral experiments. We show how residual deep neural networks (RDNNs) (Abadi et al 2016), combined with genetic algorithms (GAs), can be harnessed to classical psychophysical techniques to find different optima in high-dimensional spatiochromatic spaces This allows us to determine the best color pattern for concealment, or for signaling, in a given habitat for a given observer. A comparative study of coat colors in felids shows a correlation with ecology (Allen et al 2011), but not whether the observed patterns are the optima for the associated habitats or constrained by either the pattern-generation mechanisms or pigments available to mammals Such studies necessarily omit possible patterns that evolution has not realized because of phylogenetic or developmental constraints, and so cannot identify the influence (if any) of such constraints. Our method may be useful in the development of bespoke camouflage for specific contexts, maximizing the visibility of warning signs, or helping to reduce visual clutter due to infrastructure
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
