Some cameras will upload the photographs you snap directly to the internet; this one will do the opposite, filling its monitor with an image sourced from the web after you press its shutter. Created by the Paris-based artist and engineer Vincent Sapajou — who goes by the name Salade Tomate Oignon — Le Myope simply refuses to take new photographs of the real world, instead offering its user an existing picture similar in appearance to the desired subject.
Running on algorithms based on machine-learning and computer vision, the Raspberry Pi-based digital camera, constructed from a simple box and some electronics, uses Google images as its database. The programs, as Sapajou told Hyperallergic, essentially are “able to translate pictures into words.” So your attempt to capture a unique building mural, for instance, will instead present you with street art from another area; photographing a famous landmark, however, will likely pull up a pretty close match to the view you had in mind.
We all know that images of the objects and places we photograph most likely already exist somewhere on the internet, but Le Myope is a playful response to this routine. It introduces surprise to every shot: you’re never really sure what the device will offer you. Aiming it at a quotidian object may yield an image that transports you to a city across the world, presenting an unfamiliar view of displaced surroundings. While Sapajou describes it as “a short-sighted camera” (hence its name), I’d argue that Le Myope actually broadens our perspectives.
As engineers work to craft the sleekest, most advanced photographic tools to capture our world with the utmost precision, it’s amusing to see a growing interest in such non-camera cameras. Le Myope recalls the Camera Restricta designed by Philipp Schmitt, which prevents you from taking photographs of a place if it decides too many people have already snapped images there. Similarly, Sapajou has previously created Layer cam, a lensless camera whose screen shows popular pictures taken at your location that were uploaded to Flickr or Panoramio.
“I guess there is an appeal for low-tech, mono-purposed tools in order to [experience] another feel of, and find [new] control and understandability of the tools we use,” Sapajou said. “I wanted to have something in between — a real object, not just an app — that is both high-tech and new, but lame, totally useless, and mono-purposed as well.”
With no overly complex engineering involved, Le Myope is replicable if you’re comfortable with coding. Sapajou has shared the full DIY-guide online, so build away, and take this boxed baby out for a walk.