A photography student over in the US has developed a camera that describes a scene in text without taking an actual image. Sounds… useful, ahem.
The Descriptive Camera works like a normal camera in the sense that you point it at something and press a button. But that’s where the similarities end. Instead of getting a glorious HD reproduction of the scene, you get a print out describing what’s going on in 20 words or less.
Now, to the average Joe soap this might seem like a bit of a waste, but to the photography savvy, it could one day make categorising images a heck of a lot easier.
Creator Matt Richardson explains on his blog: “As we amass an incredible amount of photos, it becomes increasingly difficult to manage our collections.
“Imagine if descriptive metadata about each photo could be appended to the image on the fly—information about who is in each photo, what they’re doing, and their environment could become incredibly useful in being able to search, filter, and cross-reference our photo collections.
“Of course, we don’t yet have the technology that makes this a practical proposition, but the Descriptive Camera explores these possibilities.”
The camera prints out a description of the scene, instead of taking an actual photo
The camera works by sending an image captured by webcam to Amazon’s Mechanical Turk API, which uses HITs (or Human Intelligence Tasks) to figure out what’s going on in the image. So what are HITs when you’re at home?
Basically, the image gets sent to an unknown bloke sitting behind a computer who types out what he sees. The mystery man then sends this info back to the camera in ten minutes or less, and the camera prints out the description.
Check out the main image above to see what type of results the HITs bring back.
So what can we take from this? Being a HIT guy sounds like a cushy number to us.
