Digital life recorder implementing enhanced facial recognition subsystem for acquiring a face glossary data

• US 7,894,639 B2
• Filed: 11/25/2008
• Issued: 02/22/2011
• Est. Priority Date: 01/03/2008
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Wilson, John David, Grim, Clifton E. III
• Primary Examiner(s): Dwivedi; Mahesh H
• Application Number:
  US12/277,804
• Publication Number:
  US 20090175510A1
• Time in Patent Office: 819 Days
• Field of Search: 382/118, 382/115
• US Class Current: 382/118
• CPC Class Codes: G06F 16/5862 using texture

Abstract

A method and apparatus for updating a face glossary repository with face glossary data. The process identifies a source of the face glossary data in response to detecting an acquisition condition. The source may comprise a raw data queue, a first glossary from a peer digital life recorder, or a second glossary from a remote repository. The process retrieves the face glossary data from the source and stores the face glossary data from the source into the face glossary repository. The face glossary repository may include a face glossary cache and a face glossary database.

Claims

1. A computer implemented method for updating a face glossary repository with face glossary data, the computer implemented method comprising:

detecting an acquisition condition, wherein detecting the acquisition condition comprises detecting a presence of a set of face images in a raw data queue, wherein the set of face images are images recorded by a camera;

responsive to detecting the acquisition condition, identifying a source of the face glossary data;

retrieving the face glossary data from the source, wherein retrieving the face glossary data from the source comprises retrieving a voice command that refers to at least one face image in the set of face images from the source, wherein the voice command includes a timestamp that identifies a time the voice command was given;

identifying, by a processing unit, the at least one face image in the set of face images referred to by the voice command by matching the timestamp for the voice command with a timestamp that identifies a time the at least one face image was recorded by the camera to form identification information; and

storing, in the face glossary repository, the face glossary data in association with the identification information.

2. The computer implemented method of claim 1, wherein the acquisition condition is selected from a group comprising at least one of a presence of a voice command in a command queue, a preset time, a presence of the peer digital life recorder, and a presence of updated face glossary data in the remote repository, wherein the source comprises at least one of the raw data queue, a peer digital life recorder, a mobile device, and a remote repository, and wherein the face glossary repository comprises at least one of a face glossary cache and a face glossary database.

3. The computer implemented method of claim 2, wherein retrieving the face glossary data from the source further comprises:

locating the set of face images in the raw data queue identified by the voice command;

processing the set of face images to form a set of single face images, wherein each image of the set of single face images corresponds to a face in the set of face images;

performing facial metadata abstraction on the set of single face images to generate a set of processed face images; and

sending the set of processed face images to a facial recognition component to generate the face glossary data.

4. The computer implemented method of claim 2, wherein retrieving the face glossary data from the source further comprises:

responsive to the acquisition condition being the presence of a presence of the peer digital life recorder, receiving the face glossary data from the peer digital life recorder in accordance with configuration constraints, wherein the face glossary data is received from at least one of a face glossary cache of the peer digital life recorder and a face glossary database of the peer digital life recorder.

5. The computer implemented method of claim 1, wherein the voice command includes information for identifying a person in the at least one face image and wherein identifying the at least one face image in set of face images referred to by the voice command by matching the timestamp for the voice command with the timestamp that identifies the time the at least one face image was recorded by the camera to form the identification information further comprises:

assigning an identity to an image of a face in the at least one face image using the information in the voice command.

6. The computer implemented method of claim 5, wherein assigning an identity to the image includes identifying a name of the person in the at least one face image and further comprising:

performing facial metadata abstraction on the image of the face in the at least one face image to generate metadata for assigning the identity to another image of the face; and

updating the face glossary repository with the metadata.

7. A computer program product comprising:

a computer readable storage medium storing computer program code for updating a face glossary repository with face glossary data;

computer program code for detecting an acquisition condition, wherein the computer program code for detecting the acquisition condition comprises computer program code for detecting a presence of a set of face images in a raw data queue, wherein the set of face images are images recorded by a camera;

computer program code for identifying a source of the face glossary data in response to detecting the acquisition condition;

computer program code for retrieving the face glossary data from the source, wherein the computer program code for retrieving the face glossary data from the source comprises computer program code for retrieving a voice command that refers to at least one face image in the set of face images from the source, wherein the voice command includes a timestamp that identifies a time the voice command was given;

computer program code for identifying, by a processing unit, the at least one face image in the set of face images referred to by the voice command by matching the timestamp for the voice command with a timestamp that identifies a time the at least one face image was recorded by the camera to form identification information; and

computer program code for storing, in the face glossary repository, the face glossary data in association with the identification information.

8. The computer program product of claim 7, wherein the acquisition condition is selected from a group comprising at least one of a presence of a set of face images in the raw data queue, a preset time, a presence of the peer digital life recorder, and a presence of updated face glossary data in the remote repository, wherein the source comprises at least one of the raw data queue, a peer digital life recorder, a mobile device, and a remote repository, and wherein the face glossary repository comprises at least one of a face glossary cache and a face glossary database.

9. The computer program product of claim 8, wherein the computer program code for retrieving the face glossary data from the source further comprises:

computer program code for locating the set of face images in the raw data queue identified by the voice command;

computer program code for processing the set of face images to form a set of single face images, wherein each image of the set of single face images corresponds to a face in the set of face images;

computer program code for performing facial metadata abstraction on the set of single face images to generate a set of processed face images; and

computer program code for sending the set of processed face images to a facial recognition component for generating the face glossary data.

10. The computer program product of claim 8, wherein the computer program code for retrieving the face glossary data from the source further comprises:

computer program code for downloading the face glossary data from the peer digital life recorder in accordance with configuration constraints in response to the acquisition condition being the presence of the peer digital life recorder, wherein the face glossary data is downloaded from at least one of a face glossary cache of the peer digital life recorder and a face glossary database of the peer digital life recorder.

11. The computer program product of claim 7, wherein the voice command includes information for identifying a person in the at least one face image and wherein the computer program code for identifying the at least one face image in the set of face images referred to by the voice command by matching the timestamp for the voice command with the timestamp that identifies the time the at least one face image was recorded by the camera to form the identification information further comprises:

computer program code for assigning an identity to an image of a face in the at least one face image using the information in the voice command.

12. The computer program product of claim 11, wherein assigning an identity to the image includes identifying a name of the person in the at least one face image and further comprising:

computer program code for performing facial metadata abstraction on the image of the face in the at least one face image to generate metadata for assigning the identity to another image of the face; and

computer program code for updating the face glossary repository with the metadata.

13. An apparatus comprising:

a bus system;

a memory connected to the bus system, wherein the memory includes a computer usable program code;

a communications unit connected to the bus system; and

a processing unit connected to the bus system, wherein the processing unit is configured to execute the computer usable program code to;

detect an acquisition condition, wherein detecting the acquisition condition comprises detecting a presence of a set of face images in a raw data queue, wherein the set of face images are images recorded by a camera;

identify a source of face glossary data in response to detecting the acquisition condition;

retrieve the face glossary data from the source, wherein retrieving the face glossary data from the source comprises retrieving a voice command that refers to at least one face image in the set of face images from the source, wherein the voice command includes a timestamp that identifies a time the voice command was given;

identify the at least one face image in set of face images referred to by the voice command by matching the timestamp for the voice command with a timestamp that identifies a time the at least one face image was recorded by the camera to form identification information; and

store, in the face glossary repository, the face glossary data in association with the identification information.

14. The apparatus of claim 13, wherein the acquisition condition is selected from a group comprising at least one of a presence of a set of face images in the raw data queue, a preset time, a presence of the peer digital life recorder, and a presence of updated face glossary data in the remote repository, wherein the source comprises at least one of the raw data queue, a peer digital life recorder, a mobile device, and a remote repository, and wherein the face glossary repository comprises at least one of a face glossary cache and a face glossary database.

15. The apparatus of claim 14, wherein in executing the computer usable program code to retrieve the face glossary data from the source, the processing unit is further configured to execute the computer usable program code to:

locate the set of face images in the raw data queue identified by the voice command,

process the set of face images to form a set of single face images, wherein each image of the set of single face images corresponds to a face in the set of face images;

perform facial metadata abstraction on the set of single face images to generate a set of processed face images; and

send the set of processed face images to a facial recognition component for generating the face glossary data.

16. The apparatus of claim 14, wherein in executing the computer usable program code to retrieve the face glossary data from the source, the processing unit is further configured to execute the computer usable program code to:

download the face glossary data from the peer digital life recorder in accordance with configuration constraints in response to the acquisition condition being the presence of the peer digital life recorder, wherein the face glossary data is downloaded from at least one of a face glossary cache of the peer digital life recorder and a face glossary database of the peer digital life recorder.

17. The apparatus of claim 13, wherein the voice command includes information for identifying a person in the at least one face image wherein in executing the computer usable program code to identify the at least one face image in the set of face images referred to by the voice command by matching the timestamp for the voice command with the timestamp that identifies the time the at least one face image was recorded by the camera to form the identification information, the processing unit is further configured to execute the computer usable program code to:

assign an identity to an image of a face in the at least one face image using the information in the voice command.

18. The apparatus of claim 17, wherein assigning an identity to the image includes identifying a name of the person in the at least one face image and wherein the processing unit is further configured to execute the computer usable program code to:

perform facial metadata abstraction on the image of the face in the at least one face image to generate metadata for assigning the identity to another image of the face; and

update the face glossary repository with the metadata.

19. A method for updating a face glossary repository, the method comprising:

detecting a presence of a set of face images in a raw data queue, wherein the set of face images are images recorded by a camera;

retrieving a voice command that refers to at least one face image in the set of face images from the source, wherein the voice command includes a timestamp that identifies a time the voice command was given and wherein the voice command includes information for identifying a person in the at least one face image;

identifying, by a processing unit, the at least one face image in a set of face images referred to by the voice command by matching the timestamp for the voice command with a timestamp that identifies a time the at least one face image was recorded by the camera to form an identified image;

assigning an identity to an image of a face in the identified image the information in the voice command; and

storing the identity of the image in the face glossary repository.

20. The method of claim 19, wherein assigning the identity to the image includes identifying a name of the person in the identified image and further comprising:

performing facial metadata abstraction on the image of the face in the identified image to generate metadata for assigning the identity to another image of the face; and

updating the face glossary repository with the metadata.