Derrick
ლ(ಠ_ಠ )ლ
- Joined
- Nov 22, 2010
- Messages
- 5,085
- Reaction score
- 5
Imagine a world where computers can recognize images within images. A world where computers can see and understand the way we do. Imagine a computer that can look at that image you just scanned and label it for you as that trip to the coast and perhaps give you a printout of where it was taken, the time, the date and perhaps a reference as to who was there with you in the picture based on recognition from prior images you referenced.
Ever wondered how Google Images gets their image search function increasingly accurate? How is it that when you enter Springbok into Google Images it brings up images of the animal and the rugby team? You can thank (to a large degree) the workings of GWAP.
GWAP is not a hairy friend of ALF the alien that likes to eat cats famous on our television screens in the 80’s. Oh no, GWAP is our friend and he loves everything! GWAP stands for “Games With A Purpose” and it all started under the guidance of Luis von Ahn, an assistant professor in the Computer Science Department at Carnegie Mellon University.
The idea is to bring human brainpower in to play with the learning ability of computers. Think of it as a transition between human understanding and computational thinking. Computers can display but understanding is a problem for them, especially in the visual fields of video and still imaging. A computer can bring an image to your screen, but cannot tell you what is in it unless a human told it what it is. It also finds it impossible to tell different objects inside an image. Humans can do this. The skill to differentiate is something that computer scientists are working hard to fix. Intel is working with both their US and Chinese laboratories to get a working image recognition system going for household video use. Basically to allow you to fast forward to all the scenes featuring Sharon Stone for instance in Basic Instinct. So recognizing resemblances isn’t such a big issue, but knowing what it is that it is differentiating however is. Is it Sharon Stone, or is it Anna Faris it is seeing? It simply won’t know unless it is instructed as to whom (or what) it is.
The intelligence software that enables computers to learn is a work in progress and Luis von Ahn is the man that bridged another big void in this learning process for our binary friends. How do you get Joe Soap interested in teaching computers to recognize things? We ordinary folk generally value our time and when it is free time the price goes up exponentially. So how do you get folks interested in teaching computers things? You turn it into a competition between human beings.
Our need for entertainment and competative games gave birth to the phenomenon known as GWAP. Luis von Ahn created a competitive image labeling game called ESP where two human players are pitted against each other to find the same descriptor for an image. The process eliminates the chance of vandalism because it would counter the aim of the game and its process of finding the true identifier and also incidentally the most commonly recognizable identifier for a specific picture.
This was the first in now a range of games designed to be enjoyed by humans but fundamentally important to the growth we want to see in computational ability. You don’t have to sit around (idly) long to dream up something you wish a computer could help you with if only it knew how. Yes, at the end of this rabbit hole lies the holy grail of artificial intelligence par excellence and the longer we think about it’s possible implications on us the more interesting or scary it gets. Personally I believe as long as we can pull the plug all is okay, but now I am losing the plot and point of this blog.
Computers are learning! They are learning through our insatiable need for play and entertainment. Inevitably our computers are depending more and more on our human nature for their higher order development.
The best way for you to understand and learn more about what is happening in the hard core depths of computer science would be to peek a little into the classrooms of modern day computers (not the human scholars) by visiting the following recommended sites:
http://www.gwap.com
http://www.purposegames.com
http://images.google.com/imagelabeler/
Ever wondered how Google Images gets their image search function increasingly accurate? How is it that when you enter Springbok into Google Images it brings up images of the animal and the rugby team? You can thank (to a large degree) the workings of GWAP.
GWAP is not a hairy friend of ALF the alien that likes to eat cats famous on our television screens in the 80’s. Oh no, GWAP is our friend and he loves everything! GWAP stands for “Games With A Purpose” and it all started under the guidance of Luis von Ahn, an assistant professor in the Computer Science Department at Carnegie Mellon University.
The idea is to bring human brainpower in to play with the learning ability of computers. Think of it as a transition between human understanding and computational thinking. Computers can display but understanding is a problem for them, especially in the visual fields of video and still imaging. A computer can bring an image to your screen, but cannot tell you what is in it unless a human told it what it is. It also finds it impossible to tell different objects inside an image. Humans can do this. The skill to differentiate is something that computer scientists are working hard to fix. Intel is working with both their US and Chinese laboratories to get a working image recognition system going for household video use. Basically to allow you to fast forward to all the scenes featuring Sharon Stone for instance in Basic Instinct. So recognizing resemblances isn’t such a big issue, but knowing what it is that it is differentiating however is. Is it Sharon Stone, or is it Anna Faris it is seeing? It simply won’t know unless it is instructed as to whom (or what) it is.
The intelligence software that enables computers to learn is a work in progress and Luis von Ahn is the man that bridged another big void in this learning process for our binary friends. How do you get Joe Soap interested in teaching computers to recognize things? We ordinary folk generally value our time and when it is free time the price goes up exponentially. So how do you get folks interested in teaching computers things? You turn it into a competition between human beings.
Our need for entertainment and competative games gave birth to the phenomenon known as GWAP. Luis von Ahn created a competitive image labeling game called ESP where two human players are pitted against each other to find the same descriptor for an image. The process eliminates the chance of vandalism because it would counter the aim of the game and its process of finding the true identifier and also incidentally the most commonly recognizable identifier for a specific picture.
This was the first in now a range of games designed to be enjoyed by humans but fundamentally important to the growth we want to see in computational ability. You don’t have to sit around (idly) long to dream up something you wish a computer could help you with if only it knew how. Yes, at the end of this rabbit hole lies the holy grail of artificial intelligence par excellence and the longer we think about it’s possible implications on us the more interesting or scary it gets. Personally I believe as long as we can pull the plug all is okay, but now I am losing the plot and point of this blog.
Computers are learning! They are learning through our insatiable need for play and entertainment. Inevitably our computers are depending more and more on our human nature for their higher order development.
The best way for you to understand and learn more about what is happening in the hard core depths of computer science would be to peek a little into the classrooms of modern day computers (not the human scholars) by visiting the following recommended sites:
http://www.gwap.com
http://www.purposegames.com
http://images.google.com/imagelabeler/