Automatic Analysis of Individual Preferences For Attractiveness

• US 20140328536A1
• Filed: 05/02/2013
• Published: 11/06/2014
• Est. Priority Date: 05/02/2013
• Status: Active Grant

• Current Assignee: Apple Inc.
• Original Assignee: Machine Perception Technologies Inc. (Apple Inc.)
• Inventors: Whitehill, Jacob, Movellan, Javier R., Fasel, Ian
• Granted Patent: US 9,104,905 B2
• Time in Patent Office: Days
• US Class Current: 382/159
• CPC Class Codes: G06F 16/00 Information retrieval; Database structures therefor; File system structures therefor
• CPC Class Codes: G06F 16/583 using metadata automatically derived from the content
• CPC Class Codes: G06F 18/2178 based on feedback of a supervisor
• CPC Class Codes: G06V 10/7784 based on feedback from supervisors
• CPC Class Codes: G06V 40/16 Human faces, e.g. facial parts, sketches or expressions
• CPC Class Codes: G06V 40/174 Facial expression recognition

Abstract

A method facilitates selection of candidate matches for an individual from a database of potential applicants. A filter is calculated for the individual by processing images of people in conjunction with the individual'"'"'s preferences with respect to those images. Feature sets are calculated for the potential applicants by processing images of the potential applicants. The filter is then applied to the feature sets to select candidate matches for the individual.

Claims

1. A computer-implemented method for selecting candidate matches for an individual from a database of potential applicants, the method comprising a computer system performing the steps of:

calculating a filter for the individual, the filter calculated by processing images of people in conjunction with the individual'"'"'s preferences with respect to those images;

calculating feature sets for the potential applicants, the feature sets calculated by processing images of the potential applicants; and

applying the individual'"'"'s filter to the feature sets to select candidate matches for the individual.

2. The method of claim 1 wherein the step of calculating the individual'"'"'s filter comprises training the filter based on the images of people and the individual'"'"'s preferences with respect to those images.

3. The method of claim 2 wherein the step of training the filter comprises:

showing the images of people to the individual;

querying the individual for the individual'"'"'s preferences with respect to those images; and

using the images of people and the individual'"'"'s corresponding preferences as a training set to train the filter.

4. The method of claim 2 wherein the step of training the filter comprises the filter adapting over time based on the individual'"'"'s preferences to images of people.

5. The method of claim 2 wherein the step of training the filter comprises:

identifying a group of other individuals who have similar preferences as said individual; and

training the filter based on the other individuals'"'"' preferences with respect to images of people.

6. The method of claim 5 wherein the step of identifying a group of other individuals who have similar preferences as said individual comprises:

soliciting responses to questions asked of the individuals; and

identifying groups of individuals based on similarity of their responses.

7. The method of claim 2 wherein the step of training the filter comprises training the filter via machine learning.

8. The method of claim 7 wherein the step of training the filter via machine learning comprises using reinforcement learning algorithms.

9. The method of claim 1 wherein the feature sets are vectors of numbers.

10. The method of claim 1 wherein the filter is calculated by processing images of faces of people in conjunction with the individual'"'"'s preferences with respect to those images, and the step of calculating feature sets for the potential applicants is based on processing images of faces of the potential applicants.

11. The method of claim 10 wherein the filter is calculated by processing images of faces of people for facial expression in conjunction with the individual'"'"'s preferences with respect to those images, and the step of calculating feature sets for the potential applicants is based on processing images of faces of the potential applicants for facial expression.

12. The method of claim 10 wherein the filter is calculated by processing images of identifiable features in faces of people in conjunction with the individual'"'"'s preferences with respect to those images, and the step of calculating feature sets for the potential applicants is based on processing images of the identifiable features in faces of the potential applicants.

13. The method of claim 10 wherein the step of calculating feature sets for the potential applicants comprises segmenting images of faces from images of the potential applicants.

14. The method of claim 10 wherein the filter is calculated by processing images of faces of people according to a facial action coding system in conjunction with the individual'"'"'s preferences with respect to those images, and the step of calculating feature sets for the potential applicants is based on processing images of faces of the potential applicants according to the facial action coding system.

15. The method of claim 1 wherein the step of calculating feature sets for the potential applicants comprises:

segmenting the images of the potential applicants into face, body and background; and

calculating the feature sets based on the face, body and background.

16. The method of claim 1 wherein the filter is calculated by processing images of bodies of people in conjunction with the individual'"'"'s preferences with respect to those images, and the step of calculating feature sets for the potential applicants is based on processing images of bodies of the potential applicants.

17. The method of claim 1 wherein the filter is calculated by processing images of people'"'"'s clothing in conjunction with the individual'"'"'s preferences with respect to those images, and the step of calculating feature sets for the potential applicants is based on processing images of the potential applicants'"'"' clothing.

18. The method of claim 1 wherein the filter is calculated by processing backgrounds in images of people in conjunction with the individual'"'"'s preferences with respect to those images, and the step of calculating feature sets for the potential applicants is based on processing background in images of the potential applicants.

19. The method of claim 1 wherein the filter is calculated by processing images of identifiable features of people in conjunction with the individual'"'"'s preferences with respect to those images, and the step of calculating feature sets for the potential applicants is based on processing images of the identifiable features of the potential applicants.

20. The method of claim 1 wherein the filter is calculated by applying Gabor energy filters to images of faces of people in conjunction with the individual'"'"'s preferences with respect to those images, and the step of calculating feature sets for the potential applicants is based on applying Gabor energy filters to images of faces of the potential applicants.

21. The method of claim 1 wherein the filter is calculated by applying local binary pattern filters to images of faces of people in conjunction with the individual'"'"'s preferences with respect to those images, and the step of calculating feature sets for the potential applicants is based on applying local binary pattern filters to images of faces of the potential applicants.

22. The method of claim 1 wherein the database of potential applicants is a database of potential dates for the individual, and the candidate matches are candidates dates for the individual.

23. The method of claim 1 further comprising:

accessing the individual'"'"'s profile, wherein the individual'"'"'s profile contains information provided by the individual about the individual'"'"'s preferences;

accessing the potential applicants'"'"' profiles, wherein the potential applicants'"'"' profiles contain information provided by the potential applicants related to the individual'"'"'s preferences; and

further selecting candidate matches for the individual based on the information in the individual'"'"'s profile and in the potential applicants'"'"' profiles.

24. A tangible computer readable medium containing instructions that, when executed by a processor, execute a method for selecting candidate matches for an individual from a database of potential applicants, the method comprising:

calculating a filter for the individual, the filter calculated by processing images of people in conjunction with the individual'"'"'s preferences with respect to those images;

calculating feature sets for the potential applicants, the feature sets calculated by processing images of the potential applicants; and

applying the individual'"'"'s filter to the feature sets to select candidate matches for the individual.

25. A system for selecting candidate matches for an individual from a database of potential applicants, the system comprising:

means for calculating a filter for the individual, the filter calculated by processing images of people in conjunction with the individual'"'"'s preferences with respect to those images;

means for calculating feature sets for the potential applicants, the feature sets calculated by processing images of the potential applicants; and

means for applying the individual'"'"'s filter to the feature sets to select candidate matches for the individual.

26. A computer system for assisting an individual to find a date, the computer system comprising:

a database of potential dates;

a personalization module that calculates a filter for the individual, the filter calculated by processing images of people in conjunction with the individual'"'"'s preferences with respect to those images;

an analysis module that calculates feature sets for the potential dates, the feature sets calculated by processing images of the potential dates;

a match module that applies the individual'"'"'s filter to the feature sets to select candidate dates for the individual; and

a user interface to display the selected candidate dates.