Automatic suggestions to share images
Some implementations can include a computer-implemented method and/or system for automatic suggestions to share images containing people of importance to a user. The method can include determining, based on pixels of an image associated with a user account, one or more clusters associated with the image. The method can also include determining a share probability score for the image based on a probabilistic model and determining that the share probability score meets a threshold. The method can further include, in response to determining that the share probability score meets the threshold, providing a suggestion to a user associated with the user account to share the image.
1. A computer-implemented method comprising:
determining a top percentile cluster associated with a user account, wherein a first likelihood that images that include one or more people that are in the top percentile cluster are shared is higher than a second likelihood that images that do not include the one or more people that are in the top percentile cluster are shared;
determining that a face in a particular image corresponds to the one or more people that are in the top percentile cluster;
determining a share probability score that indicates a probability the particular image will be shared based on a probabilistic model, wherein the probabilistic model is based on an account size of the user account and the top percentile cluster;
determining that the share probability score meets a threshold value; and
in response to determining that the share probability score meets the threshold value, providing a suggestion to a user associated with the user account to share the particular image.
2. The method of claim 1 , wherein the account size includes a total number of images.
3. The method of claim 1 , wherein the probabilistic model is further based on one or more of a total number of images associated with the user account, a number of clusters associated with the user account, or a number of clusters that have a name label provided by the user.
4. The method of claim 1 , wherein the probabilistic model is further based on a probability that the user shares any image, a probability that a shared image includes at least one person that is in the one or more people that are in the top percentile cluster, and a probability that any image includes at least one person that is in the one or more people that are in the top percentile cluster.
5. The method of claim 1 , wherein:
the top percentile cluster is a person cluster with a ranking that meets a ranking percentile threshold; and
the ranking is based on at least one of face quality, a number of images associated with the user account, a recency of images in a set of clusters, a count of time periods of the set of clusters, and whether a person in an image was assigned a name label by the user.
6. The method of claim 1 , wherein the face is a dominant face in the particular image, and further comprising identifying that the face is the dominant face in the particular image by applying a facial recognition technique to the particular image.
7. The method of claim 1 , further comprising ranking a set of person clusters, wherein determining the top percentile cluster is based on identifying a top ranked person cluster from the set of person clusters.
8. A system comprising:
one or more processors; and
a memory that stores instructions that, when executed by the one or more processors cause the one or more processors to perform operations comprising:
determining a top percentile cluster associated with a user account, wherein a first likelihood that images that include one or more people that are in the top percentile cluster are shared is higher than a second likelihood that images that do not include the one or more people that are in the top percentile cluster are shared;
determining that a face in a particular image corresponds to the one or more people that are in the top percentile cluster;
determining a share probability score that indicates a probability the particular image will be shared based on a probabilistic model, wherein the probabilistic model is based on an account size of the user account and the top percentile cluster;
determining that the share probability score meets a threshold value; and
in response to determining that the share probability score meets the threshold value, providing a suggestion to a user associated with the user account to share the particular image.
9. The system of claim 8 , wherein the account size includes a total number of images.
10. The system of claim 8 , wherein the probabilistic model is further based on one or more of a total number of images associated with the user account, a number of clusters associated with the user account, or a number of clusters that have a name label provided by the user.
11. The system of claim 8 , wherein the probabilistic model is further based on a probability that the user shares any image, a probability that a shared image includes at least one person that is in the one or more people that are in the top percentile cluster, and a probability that any image includes at least one person that is in the one or more people that are in the top percentile cluster.
12. The system of claim 8 , wherein:
the top percentile cluster is a person cluster with a ranking that meets a ranking percentile threshold; and
the ranking is based on at least one of face quality, a number of images associated with the user account, a recency of images in a set of clusters, a count of time periods of the set of clusters, and whether a person in an image was assigned a name label by the user.
13. The system of claim 8 , wherein the face is a dominant face in the particular image, and further comprising identifying that the face is the dominant face in the particular image by applying a facial recognition technique to the particular image.
14. The system of claim 8 , wherein the operations further comprise ranking a set of person clusters, wherein determining the top percentile cluster is based on identifying a top ranked person cluster from the set of person clusters.
15. A non-transitory computer-storage medium encoded with a computer program, the computer program comprising instructions that, when executed by one or more computers, cause the one or more computers generate a video by performing operations comprising:
determining a top percentile cluster associated with a user account, wherein a first likelihood that images that include one or more people that are in the top percentile cluster are shared is higher than a second likelihood that images that do not include the one or more people that are in the top percentile cluster are shared;
determining that a face in a particular image corresponds to the one or more people that are in the top percentile cluster;
determining a share probability score that indicates a probability the particular image will be shared based on a probabilistic model, wherein the probabilistic model is based on an account size of the user account and the top percentile cluster;
determining that the share probability score meets a threshold value; and
in response to determining that the share probability score meets the threshold value, providing a suggestion to a user associated with the user account to share the particular image.
16. The computer-storage medium of claim 15 , wherein the account size includes a total number of images.
17. The computer-storage medium of claim 15 , wherein the probabilistic model is further based on one or more of a total number of images associated with the user account, a number of clusters associated with the user account, or a number of clusters that have a name label provided by the user.
18. The computer-storage medium of claim 15 , wherein the probabilistic model is further based on a probability that the user shares any image, a probability that a shared image includes at least one person that is in the one or more people that are in the top percentile cluster, and a probability that any image includes at least one person that is in the one or more people that are in the top percentile cluster.
19. The computer-storage medium of claim 15 , wherein:
the top percentile cluster is a person cluster with a ranking that meets a ranking percentile threshold; and
the ranking is based on at least one of face quality, a number of images associated with the user account, a recency of images in a set of clusters, a count of time periods of the set of clusters, and whether a person in an image was assigned a name label by the user.
20. The computer-storage of claim 15 , wherein the face is a dominant face in the particular image, and further comprising identifying that the face is the dominant face in the particular image by applying a facial recognition technique to the particular image.