Army seeking psychologically inspired object recognition system

Yeah, as in, it actually wants a "psychologically inspired object recognition system." What's that, you ask? It's giving robots and mechanical creatures the ability to see objects the way humans do and make reasonable judgments based on those sights. Essentially, the military would love to see bots have something similar to spatial memory, which would enable 'em to "mentally rotate objects in order to match the object to different representations." When looking at the main objective of this here endeavor, however, we can't help but have mixed feelings. We're kosher with increasing "robotic control," but creating "exponential expansion of robotic capabilities and intelligence" might not be the smartest thing to do in the long run.

[Via Wired, image courtesy of ACM]

[SuperSexyErik @ Jul 19th 2008 12:21AM]

iRobot
Get on it apple :)

[JMMGoalster @ Jul 19th 2008 12:55AM]

hmm...you do know that iRobot is a legit company right? Not owned by Apple?

[Matthewmongan @ Jul 19th 2008 1:05AM]

It is seriously odd that iRobot (of Roomba fame) calls itself iRobot. Visit the sci fi section of your public library much, anyone?

[Matthewmongan @ Jul 19th 2008 12:59AM]

Johnny 5?

[feffrey @ Jul 19th 2008 1:07AM]

Is Alive!

[Artie Lange @ Jul 19th 2008 12:48AM]

One wonders what DARPA's actual goal is when they put out these RFPs. SOP is to compartmentalize requests so that foreign militaries can't figure out the big picture. So what does DARPA want to do with this? Or rather, what does it want to give this capability to?

[E71 @ Jul 19th 2008 5:14AM]

No disassemble Number Five!

[Hello Moto @ Jul 19th 2008 12:29PM]

It allows the robot to rotate any form it sees in 3-d dimensions to see if the object is in fact a human. Right now if you contorted into a funny shape they might not see you, but with this advanced programming? You’re good as gone even if you do dislocate all your joints to bend them all in the wrong direction!

[mustardman @ Jul 19th 2008 12:54AM]

douche

[waiownsyou @ Jul 19th 2008 12:57AM]

Your mom is better than my girlfriend but you don't see me telling everyone that. Do you?

Settle down.

[Coldflame @ Jul 19th 2008 1:01AM]

Skynet much?
Somebody find Miles Dyson

[bolezhinkov @ Jul 19th 2008 1:25AM]

I for one, welcome our . . .

in related news: robots such as that would be awesome and would take over the world.

[(01) @ Jul 19th 2008 2:00AM]

Good thing I just finished watching Terminator and this doesn't terrify me at all...Oh wait.

[gilmorojamiro @ Jul 19th 2008 2:38AM]

A lot of wires

[BadBuddy234 @ Jul 19th 2008 2:42AM]

nice pic with those wires...

[Colin @ Jul 19th 2008 6:53AM]

NUMBA JOHNNY FIVE.

[Phoenixfury @ Jul 19th 2008 8:15AM]

I'm afraid to say this but.. That thing looks like a cross between Johnny 5, the terminator, and Jar Jar Binks! Perhaps a terminator somehow teleported to a time long long ago, to a galaxy far far away.. It's sole mission.. To kill Jar Jar Binks.. So unfortunately it had to look like him.. Upon terminating Jar Jar, it is transitioning itself. Johnny 5 is next.. (gasp!)

[FiyaWiya @ Jul 19th 2008 10:48AM]

"mentally rotate objects in order to match the object to different representations."

As in, "Are you Sarah Connor?"

[Sam Stone @ Jul 19th 2008 11:27AM]

Wall-E looks like he hit puberty...

[Dirk @ Jul 19th 2008 2:14PM]

So long as they don't send one of those emotional robots into the field by accident, goading the angry gun-toting enemy with remarks of 'why so sad?' and harassing them with cuddles.

[Klintor @ Jan 5th 2009 12:34PM]

DUDE! How come nobody's named the picture yet?! I remember that guy, Cog. The first robot to "learn" solely by interaction and experimentation. It's been, like, 9 years!

[tsishkou @ Jul 20th 2008 4:48AM]

have a look at

http://www.semantic-robot-vision-challenge.org/

The Semantic Robot Vision Challenge (abbreviated SRVC) is a research competition that is designed to push the state of the art in image understanding and automatic acquisition of knowledge from large unstructured databases of images (such as those generally found on the web).

In this competition, fully autonomous robots receive a text list of objects that they are to find. They use the web to automatically find image examples of those objects in order to learn visual models. These visual models are then used to identify the objects in the robot's cameras.