Welcome to WebmasterWorld Guest from 22.214.171.124
How does it work?
You'll be randomly paired with a partner who's online and using the feature. Over a 90-second period, you and your partner will be shown the same set of images and asked to provide as many labels as possible to describe each image you see. When your label matches your partner's label, you'll earn some points and move on to the next image until time runs out. After time expires, you can explore the images you've seen and the websites where those images were found. And we'll show you the points you've earned throughout the session.
<edit reason: limit on length of quotes.
See Terms of Service [webmasterworld.com]
[edited by: tedster at 8:01 pm (utc) on Sep. 1, 2006]
Also a match apparently has to be exact, which rewards generic tags over more specific ones. For example I got a picture of Pope Benedict XVI but it was only when I finally got around to trying "pope" did I score a match, even though I'd included the word "pope" in earlier attempts. So if you actually recognize the specifics of a picture as I did in several cases you're probably at a disadvantage. Google still gets the benefit of my input but once again I don't get the cheese.
P.S. They can make a “learn how to type” game out of this – this way they get their answers and I would eventually learn how to type :c)
Here's how we test the weakness of it, and the strength of WebmasterWorld.
The only "news story" about this in Google results is one listed as appearing 32 minutes ago in ITnews;
"...Building a database using information provided by volunteers has its risks: campaign groups or pranksters might influence or pollute the raw data by associating an insulting term with the image of a political candidate..."
If we all play from 10:00pm -10:15pm (EST), and use the label "snark pic" for every picture, we'll know if we are playing with someone else who read this post, and how easy it is to beat.
I am sure G will appreciate us "contributing" the testing.
The ESP Game is a two-player game. Each time you play you are randomly paired with another player whose identity you don't know. You can't communicate with your partner, and the only thing you have in common with them is that you can both see the same image. The goal is to guess what your partner is typing on each image. Once you both type the same word(s), you get a new image.
also see: [web.archive.org...]
Yet another Google rip-off idea.
I'll leave it to others to worry about whether the "ESP Game" predates Carnegie Mellon's implementation or not, or whether it makes any difference either way.
did anyone notice their WebmasterWorld nick gets them logged in as such?
anyone got privacy concerns about how this was integrated with WebmasterWorld?
next time, please give us fair warning and let us have a choice.
...somehow my gmail account (which includes sitemap and adwords) appeaers to be now associated with my 'anonymous' nick here.
good thing I'm not trying to hide anything!
I played it earlier and it worked fine, just tried it now and game appears broken --- long delays to get a partner, (even though it supposedly has an AI mode that lets you play a two player game and defy the time space continuem even of there are no other players), game doesn't quit after both players agree to pass, (clock continues to tick after passing - does not move to next image)... wassup?
I deleted cookies and reloaded browser to make sure it wasn't holding a grudge against me for earlier comments --- I'll try this;
Google is great,
Google is Great,
Google is Great
...now can I play more?
or they will make a robot that can go into a room and tell you what items are in that room in seconds.
this will help them with search too, googlebot visits a webpage and automatically knows what images are used on a webpage. no more making up ALT tags.
plus it will help desktop search, users searching for a 'dog' photo on their desktop will be able to find just that with the help of google's image recognition.
Most people use general words like person, people, group, man, boy, girl etc.
The system will automatically add words to the "off-limit" list. I've seen the same image (a particularly difficult one to label) come up in rotation three times with a different set of "off-limit" words. Each time it's clear that the system is getting the same results over and over again and is trying to force varied input.
I'm sure the "off-limit" list is also used to accomplish other goals, but it's clear it can solve the problem of everyone using the same generic keywords easily.
After posting that, chowmein pointed out a video clip which was very interesting and disclosed some of the security measure built into the labeler---
Pre-determined test images for which google knows what the labels should be are presented to users early in the game, (ie- a bird, a dog, a tree). If one of the players gives labels like "snark" on these, G knows the user is gaming the game... If both of the remote users (who are supposedly unknown to each other) provide the same (false) label, it tells the system it is being gamed by both players.
They did extensive testing with the software under the name "The ISP Game" prior to it's use as the Google Image Labeler.
If you see any text on the screen just mention it for a match so that could help Google.
If you try it reduce your screen resolution to a minimum it helps.