Cuthill and Stevens revived interest in disruptive camouflage, but the first real insights into just how it works came only last year. Richard Webster at Carleton University in Ottawa, Canada, asked volunteers to search for virtual moths on a computer screen while an eye-tracker monitored their gaze. “We could almost get inside people’s eyes,” he says. He found that the more patches moths had on their edges, the more often volunteers failed to notice them, and they needed to fixate their gaze on them for longer to have any chance of spotting them. The eyetracking vindicated Thayer again: by breaking up an animal’s outline, disruptive camouflage does impair a predator’s ability to spot its prey. Although instructive, the experiment had an obvious shortcoming: humans do not prey on moths, let alone computer-generated ones. To test whether disruptive colouring fools its intended audience, Stevens has started field trials. In Zambia and South Africa, his team is studying groundnesting birds that rely on disruptive camouflage, including nightjars and plovers. His team measures the patterns on the birds’ feathers to quantify how well hidden they are in their environment. They also track the birds’ survival to determine how effectively they evade predators. Nightjars and plovers are difficult to spot in the first place, so the researchers have employed sharpsighted local guides to help find them. This raises the question of whether predators, like the guides, might be less easily fooled by disruptive markings as they become more familiar with them. Last year, Stevens and his team found that people do gradually get better at spotting virtual moths, especially if they see several at the same time. He suspects that the volunteers learn to stop the futile search for outlines, and instead start scanning for the highcontrast markings. Whether non-human predators adopt the same tactic is hard to say. They may not even see camouflage markings in the same way that we do. But if predators can learn to see through disruptive camouflage, it would suggest that this concealment strategy is more likely to evolve in prey that face short-lived or generalist predators than long-lived or specialist ones. Another open question is whether one disruptive pattern might work in a variety of environments. Webster found that camouflage can fool humans even when the colours do not match the background, provided they are not too garish. “A jester’s costume is highly disruptive, but he’s always going to stand out,” he says. Perhaps that explains the British army’s recent decision to replace its long-standing woodland and desert camouflage patterns with a single “multi-terrain pattern” that includes green, brown
During both world wars, cubism was camouflage
and sandy yellow, following field trials indicating this print was better at concealing soldiers. “Cynics may say this is just a cost-cutting exercise,” says Webster. “But maybe they’ve worked out how to get the most out of disruptive colouration.” In fact, the military has a long history of using conspicuous patterns to fool observers. During both world wars, several US and British ships were painted with striking black-and-white geometric patterns. Some looked like floating checkerboards, others like Cubist zebras. Rather than hiding the vessels, these gaudy designs, known as “dazzle camouflage” in the UK and
razzle-dazzle” in the US, were supposed to make it harder for the enemy to judge a ship’s speed, size and bearing. There was scant evidence to support the idea, so Cuthill decided to test it. He found that people consistently underestimated the speed of on-screen dazzle patterns, such as checks and zigzags, by around 7 per cent – but only when the patterns moved quickly. Large warships are probably too slow to benefit from this “motion dazzle” illusion, he concludes, but it could allow a speeding Land Rover to evade enemy fire. Conspicuous markings might, likewise, help a fleeing zebra elude a pursuing lion
Bold make-up and bizarre hairstyles might thwart facial recognition” The army is not alone in wanting to exploit Thayer’s insights. Governments’ efforts to keep an eye on their citizens are changing rapidly, and this year the Janus programme, run by the US intelligence community, will start collecting photographs from social media websites and public video feeds, to identify faces in the images by using algorithms that match them to those on existing web profiles. Face recognition could also feature in drones, which are set to become a common feature outside war zones. Alarmed by the steady rise in surveillance, New York-based artist Adam Harvey aims to thwart facial recognition technologies by being conspicuous. The bold make-up and bizarre hairstyle showcased in his CV Dazzle project are certainly eccentric; whether they can foil an AI system is another matter. But it’s tempting to think that Thayer would have approved. ■
Aucun commentaire:
Enregistrer un commentaire