METHODS FOR TAGGING PERSON IDENTIFICATION INFORMATION TO DIGITAL DATA AND RECOMMENDING ADDITIONAL TAG BY USING DECISION FUSION

• US 20090256678A1
• Filed: 10/31/2006
• Published: 10/15/2009
• Est. Priority Date: 08/17/2006
• Status: Active Grant

• Current Assignee: Intel Corporation
• Original Assignee: Olaworks, Inc.
• Inventors: Ryu, Jung-hee
• Granted Patent: US 7,634,115 B2
• Time in Patent Office: Days
• US Class Current: 340/5.830
• CPC Class Codes: G06F 16/58 Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
• CPC Class Codes: G06V 40/172 Classification, e.g. identification
• CPC Class Codes: G06V 40/179 metadata assisted face recognition

Abstract

Person identification information is extracted from digital data with a high accuracy and additional tag is recommended by adopting a decision fusion. In other words, the person identification information is acquired by using various additional information to tag the same to the digital data automatically as a person tag, in a digital data management system in which the person identification information is extracted from the digital data by referring to its attributes and contents; and specific digital data having specific tags which are same as those attached to newly created digital data are found out and then candidate tags which are attached to the specific digital data except the specific tags are provided to a monitor in order for a user to choose one or more proper tags from the candidate tags, which are desired to be attached to the newly created digital data additionally, by using the decision fusion.

Claims

1. A method for tagging identification information of a first person to a first digital data created by using a digital equipment in a digital data management system, wherein the digital data management system includes both identification information and life pattern information of a user who creates a plurality of digital data by using the digital equipment, and the first digital data includes an image of the first person, the method comprising the steps of:

(a) identifying the first person by using at least one out of additional information of the first person; and

(b) automatically attaching the identification information of the first person to the first digital data,

wherein the additional information includes at least one of information;

information which indicates that if a first digital data group includes an image of a second person, a probability that the first person is determined to be the second person is increased, the first digital data group being comprised of the first digital data and other digital data distributed in a same cluster as that including the first digital data,

information which indicates that if the first digital data is created while two or more people have a meeting, a probability that the image of the first person is of a transmitter or a receiver of communication made before the meeting is increased,

information which indicates that if an image of a third person is included m times in a second digital data group, a probability that the first person is the same as the third person increases as the number of the m increases, the second digital data group being comprised of the first digital data and other digital data, the m being a natural number,

information associated with the first person acquired by using at least one of a time when the first digital data is created and a place where the first digital data is created as well as the life pattern information of the user of the digital equipment,

information which indicates that a probability of the first person being determined to be a specific person is varied according to a position and a size of a face of the first person on the image of the first person,

information which indicates that if the digital equipment is a dual camera and the first digital data is created through a lens of the dual camera directed toward the user, a probability that the first person is determined to be the user of the digital equipment is increased, and

information which indicates that if the digital equipment is the dual camera and the first digital data is created through a lens of the dual camera directed toward the opposite side of the user, a probability that the first person is determined to be the user of the digital equipment is decreased.

2. The method of claim 1, wherein the step (a) includes a step of combining the additional information by a decision fusion, the decision fusing being a general term of a process for making a decision on a specific issue from multiple cues and a priori probability.

3. The method of claim 2, wherein the decision fusion includes Bayesian inference or ad-hoc analysis.

4. The method of claim 3, wherein the step (a) further includes a step of creating a top-n list, in which candidates having the top n probability of being determined to be the first person are included, based on the result of the combination of the additional information through the Bayesian inference or the ad-hoc analysis.

5. The method of claim 4, wherein the top-n list is updated in real-time through the Bayesian inference or the ad-hoc analysis.

6. The method of claim 4 or 5, wherein the step (b) includes a step of determining the first person to be a person having the highest probability among the candidates of top-n list.

7. The method of claim 1, wherein the communication is at least one of e-mail communication, wire or wireless telephone communication, SMS message communication, and messenger communication.

8. The method of claim 1, wherein the information on the receiver or the transmitter of the communication made before the meeting is acquired from a communication device used for the meeting.

9. The method of claim 1, wherein the information on the receiver or the transmitter of the communication made before the meeting is acquired by recognizing contents of the communication.

10. The method of claim 1, wherein the information associated with the first person acquired by using at least one of the time when the first digital data is created and the place where the first digital data is created as well as the life pattern information of the user of the digital equipment is:

information which indicates that a probability of the first person being determined to be a family member or a friend of the user is increased if the time corresponds to a holiday, and

information which indicates that a probability of the first person being determined to be a fellow worker of the user is increased if the time corresponds to a weekday.

11. The method of claim 1, wherein the information associated with the first person acquired by using at least one of the time when the first digital data is created and the place where the first digital data is created as well as the life pattern information of the user of the digital equipment is:

information which indicates that a probability of the first person being determined to be a friend or a fellow worker of the user is increased if the place is a drinking house or a bar, and

information which indicates that a probability of the first person being determined to be a family member of the user is increased if the place is home.

12. The method of claim 11, wherein the place is perceived by referring to tags attached to the first digital data.

13. The method of claim 1, wherein if the first digital data further includes an image of a fourth person, the information which indicates that a probability of the first person being determined to be a specific person is varied according to a position and a size of a face of the first person on the image of the first person is:

information which indicates that a probability of the first person being determined to be the specific person is decreased if a face of the fourth person is positioned at the center of the whole image of the first digital data and the face of the fourth person is remarkably larger than that of the first person.

14. The method of claim 1, wherein if the first digital data further includes an image of a fifth person, the information which indicates a probability of the first person being determined to be a specific person is varied according to a position and a size of a face of the first person on the image of the first person is:

information which indicates that a probability of the first person being determined to be a friend of a fifth person is increased if a size of the face of the first person is similar to that of a face of the fifth person.

15. The method of claim 14, wherein the fifth person is identified by referring to tags attached to the first digital data.

16. A method for assigning one or more second tags to a first digital data additionally, the second tags being different from a first tag which has been attached to the first digital data, wherein a specific user creates the first digital data, including at least one of images of a specific person P or a specific object O at a specific space S in a specific time zone T of a day, by using a digital equipment,

wherein at least one of information on the specific space S, the specific person P, the specific object O, and the specific time zone T has been attached to the first digital data as the first tag,

wherein the first digital data which is newly created and a plurality of existing digital data are stored in a database included in a digital data management system, the method comprising the steps of:

(a) searching identical_tag_on_digital_data having the same tag as the first tag from the existing digital data;

(b) calculating how frequently each of residual tags, excluding the same tags as the first tag, from all the tags assigned to the identical_tag_on_digital_data, is assigned to the identical_tag_on_digital_data;

(c) displaying candidate tags, having the top N frequency among the residual tags, on a screen of the digital equipment;

(d) selecting at least one of the candidate tags by the user; and

(e) assigning the selected candidate tags to the first digital data as the second tags of the first digital data.

17. The method of claim 16, wherein the frequency is calculated by a decision fusion, the decision fusion being a general term of a process for making a decision on a specific issue from multiple cues and a priori probability.

18. The method of claim 17, wherein the decision fusion includes Bayesian inference or ad-hoc analysis.

19. The method of claim 18, wherein the candidate tags are updated in real-time.