Most cameras these days are pretty complicated with settings for just about anything you can think of. The Descriptive Camera, on the other hand, is pretty simple: point the camera, press the shutter and an image comes out like a Polaroid. This camera, however, instantly produces an image that is a text description of the content — no photo at all.
The camera consists of a USB webcam, a shutter button, a small printer and an Ethernet connection. Once the user 'takes a photo', the image is transmitted via the Ethernet cable to Amazon’s Mechanical Turk API, where workers completing 'human intelligence tasks' (HITs) take a look at the image and then describe what they see. Once a description has been assigned to the image (through a combination of HIT responses), the thermal printer on the camera prints out the text.
Take a photo of a cupboard for example — the text description comes out with, 'looks like a cupboard which is ugly and old having name plates on it with a study lamp attached to it'. A photo of a building, meanwhile, yields, 'this is a faded picture of a dilapidated building. It seems to be run down and in need of repairs.'
The project was created by Matt Richardson as part of New York University's Telecommunications course with the hope that one day it could be used for managing photography collections. He explains, ' I was picturing a time in which cameras could possibly capture more useful information that can then be searched, cross-referenced and sorted.'
Check it out here.
No comments:
Post a Comment