Systems and methods for performing image clustering
AssigneeCyberlink Corp.
InventorsShih-Min Tang, Chi Ming Lee
Life of this patent
Estimating the term…
Abstract
One embodiment, among others, is a method for clustering a plurality of images, wherein the plurality of images comprises faces of a plurality of individuals. The method comprises arranging the plurality of images associated with a plurality of individuals into a plurality of subgroups for each individual based on time stamps associated with the plurality of images, wherein the plurality of images are arranged according to increments of a time interval. The method further comprises determining whether adjacent subgroups are correlated and forming groups comprising correlated subgroups. Based on correlations between adjacent groups, the groups are associated with a particular individual.
Claims21
A method for clustering a plurality of images, wherein the plurality of images comprises faces of a plurality of individuals, the method comprising:
- arranging the plurality of images associated with a plurality of individuals into at least two groups using facial clustering, wherein the first group is clustered based on a first temporal information comprising a first start time and first end time, and the second group is clustered based on a second temporal information comprising a second start time and second end time, wherein the second start time is between the first start time and the first end time;
- determining whether any of the at least two of groups include at least one overlapping image;
- and
- in response to determining that any groups contain at least one overlapping image, associating the overlapping groups with one individual.
A method for clustering a plurality of images, wherein the plurality of images comprise faces of a plurality of individuals, the method comprising:
- arranging the plurality of images associated with a plurality of individuals into a plurality of subgroups for each individual based on time stamps associated with the plurality of images, wherein the plurality of images are arranged according to increments of a time interval;
- determining whether adjacent subgroups are correlated;
- forming groups comprising correlated subgroups;
- and
- based on correlations between adjacent groups, associating the groups with a particular individual.
A program stored on a computer readable medium for clustering images of individuals, the program comprising:
- logic for arranging the plurality of images associated with a plurality of individuals into at least two groups using facial clustering, wherein the first group is clustered based on a first time interval and the second group is clustered based on a second time interval;
- logic for determining quantities of images associated with the first and second time interval and based on the image quantities, setting the first and second time interval;
- analysis logic for determining correlation between the at least two groups;
- and
- logic for associating correlated groups with one individual based on whether the at least two groups are correlated.
Dependencies are inferred from the claim text.