Method for labeling images through a computer game

• US 7,980,953 B2
• Filed: 06/24/2004
• Issued: 07/19/2011
• Est. Priority Date: 07/01/2003
• Status: Active Grant

• Current Assignee: Carnegie Mellon University
• Original Assignee: Carnegie Mellon University
• Inventors: von Ahn Arellano, Luis
• Primary Examiner(s): McClellan; James S
• Application Number:
  US10/875,913
• Publication Number:
  US 20050014118A1
• Time in Patent Office: 2,581 Days
• Field of Search: 463/42, 463/30, 463/31
• US Class Current: 463/42
• CPC Class Codes: A63F 13/12 involving interaction between a plurality of game devices, e.g. transmisison or distribution systems protocols for networked virtual reality, networked games H04L29/06034
• CPC Class Codes: A63F 13/30 Interconnection arrangements between game servers and game devices; Interconnection arrangements between game devices; Interconnection arrangements between game servers
• CPC Class Codes: A63F 13/335 using Internet
• CPC Class Codes: A63F 13/795 for finding other players; for building a team; for providing a buddy list
• CPC Class Codes: A63F 13/80 Special adaptations for executing a specific game genre or game mode
• CPC Class Codes: A63F 2300/5566 by matching opponents or finding partners to build a team, e.g. by skill level, geographical area, background, play style
• CPC Class Codes: A63F 2300/8094 Unusual game types, e.g. virtual cooking
• CPC Class Codes: G06F 16/583 using metadata automatically derived from the content

Abstract

A methodology to determine the contents of an image uses an online game that is played by a large number of people at once. Each player may be randomly assigned a partner to form a pair. Partners do not know each other'"'"'s identity and cannot communicate with one another. Players cannot see what their partners are typing, and the only thing the two partners have in common is an image that they can both see. One object of the game is to have each player type exactly what his or her partner is typing or has already typed. Once both players type the same string of words or letters, they receive a new image (the players are not required to type the string at the same time). The string upon which the two players agree is used as a label for that image. When trying to agree on a word, players may not be able to type certain specific strings of characters or “forbidden words”. When humans play the electronic game (online or over the Internet), they help computers determine the contents of images. If played by a large number of people, this game could provide a resource capable of classifying a vast portion of images on the World Wide Web. Because of the rules governing abstracts, this abstract should not be used to construe the claims.

Claims

1. A method comprising:

selecting at least two participants from a plurality of participants and grouping the at least two participants;

presenting a first image to the at least two selected participants;

receiving from each of the at least two selected participants, a content-identifying term for said first image;

determining, via a computer system, if the content-identifying term received from the at least two selected participants match; and

responsive to determining that the content-identifying terms match and that the matching content-identifying terms are not included in a list of forbidden terms, storing the matching content-identifying term as a label for the first image, and adding the matching content-identifying term to the list of forbidden terms.

2. The method of claim 1, wherein said selecting is performed randomly.

3. The method of claim 1, wherein said content-identifying term is a written text input electronically by each of the at least two selected participants.

4. The method of claim 1, wherein said selecting the at least two participants includes:

pairing two of said plurality of participants without disclosing an identity of one participant to the other and without disclosing the content identifying term provided by one participant to the other participant.

5. The method of claim 1, wherein said presenting the first image includes:

accessing a database storing a plurality of images; and

selecting one of said plurality of images to be presented as said first image to each said selected participant.

6. The method of claim 5, wherein selecting one of said plurality of images includes randomly selecting said one of said plurality of images in said database as said first image.

7. The method of claim 1, further comprising:

storing said content-identifying terms of said first image provided by each said selected participant;

comparing the content-identifying terms received from each said selected participant for similarity as each content-identifying term is received; and

presenting a second image to each said selected participant when the content-identifying terms of said first image received from each said selected participant match.

8. The method of claim 7, wherein presenting said second image includes removing said first image from the view of each said selected participant.

9. The method of claim 1,

wherein the matching content-identifying term is stored responsive to the matching content-identifying term further not being

a textual variation related to a forbidden term in the list of forbidden terms.

10. A method, comprising:

selecting two participants from said plurality of participants and pairing the two selected participants;

presenting a first image to said selected two participants so as to enable both of said selected participants to view said first image;

receiving from each of said two selected participants at least one content-identifying term for said first image; and

responsive to determining, via a computer system, that the content-identifying terms match, storing the content-identifying term as a label for the first image and presenting a second image to the selected two participants so as to enable both of the selected participants to view the second image.

11. The method of claim 10, wherein said providing includes one of the following:

offering an online electronic game over a communication network on a subscription-basis to said plurality of participants; and

charging two or more of said plurality of participants on a per-game basis for said online electronic game.

12. The method of claim 11, wherein said communication network is the Internet.

13. The method of claim 12, wherein said providing includes:

hosting said online electronic game at a web site on the Internet; and

allowing each said participant to access said website using a computing unit.

14. The method of claim 10, wherein said selecting includes:

pairing two of said plurality of participants without disclosing an identity of one participants to the other and without disclosing the content-identifying term provided by one participant to the other participant.

15. The method of claim 10, wherein said presenting includes:

accessing a database storing a plurality of images over a communication network; and

selecting one of said plurality of images to be presented as said first image to each of said two selected participants.

16. The method of claim 15, wherein selecting one of said plurality of images includes randomly selecting said one of said plurality of images.

17. The method of claim 10, wherein said content-identifying term is a written text.

18. The method of claim 10, further comprising:

presenting the second image to the first and second participants responsive to a predetermined time period elapsing without receiving the matching content-identifying terms from said two selected participants.

19. The method of claim 18, wherein presenting said second image includes removing said first image from the view of both of said selected participants.

20. The method of claim 10,

wherein the content-identifying terms are not stored as a label for the first image responsive to the content identifying term comprising at least one of:

a word that has already been used as part of said corresponding content-identifying term for said first image;

a textual variation related to said word; and

a forbidden term.

21. A method, comprising:

selecting two participants from said plurality of participants and pairing the two selected participants;

presenting a first image to the selected participants; and

receiving content-identifying terms describing the image from each of the two selected participants;

responsive to the content-identifying terms received from the two selected participants matching and both of the content-identifying terms being received within a predetermined time period, storing the matching content-identifying term in a data storage medium as a label for the first image; and

responsive to the predetermined time period expiring prior to the two selected participants providing the matching content-identifying term, presenting a second image to the selected participants.

22. The method of claim 21, wherein said requesting includes:

inviting each of said two selected participants to input a term for said image within a predetermined time.

23. A method comprising:

selecting two participants from a plurality of participants;

presenting a first image to each of said two selected participants;

receiving from each of said two selected participants at least one content-identifying term for said first image;

removing said first image from the view of both of said selected participants, via a computer system, and presenting a second image to said selected two participants upon occurrence of one of the following;

receiving from each of the two selected participants a matching content-identifying for the first image, wherein the matching content-identifying term is not included in a list of forbidden terms; and

a predetermined time period elapsing prior to receiving the matching content-identifying term not included in the list of forbidden terms from each of said two selected participants.

24. A data storage medium containing program instructions, which, when executed by a processor, cause said processor to perform the following:

select two participants from said plurality of participants and pairing the two selected participants;

present a first image to the selected participants; and

receive a content-identifying term describing the image from each of the two selected participants;

responsive to the content-identifying terms received from the two selected participants matching and both of the content-identifying terms being received within a predetermined time period, store the matching content-identifying term as a label for the first image; and

responsive to the predetermined time period expiring prior to the two selected participants providing the matching content-identifying term, present a second image to the selected participants.

25. The data storage medium of claim 24, wherein said program instructions, upon execution, cause said processor to further perform the following:

invite each of said two selected participants to input an identical term for said image within a predetermined time.

26. A method comprising:

selecting a first participant and a second participant from a plurality of participants and pairing the first and second participant;

presenting an image to each of said first and said second participants;

receiving a plurality of terms describing the image from each of said first and said second participants;

recording the received plurality of terms describing the image from each of said first and said second participants;

presenting said image to a third participant from said plurality of participants;

receiving a content-identifying term for the image from said third participant; and

supplying to said third participant, via a computer system, a different recorded term from said recorded plurality of terms in response to each content-identifying term for said image supplied by said third participant.

27. The method of claim 26, further comprising:

responsive to the content-identifying term from the third participant matching a recorded term from the recorded plurality of terms, presenting a second image to the third participant and storing the matching content-identifying term as a label for the first image.