Location-Based Facial Recognition on Online Social Networks

Abstract

In one embodiment, a computing system may access an image associated with an online social network, where the image is associated with a first user and portrays at least a first person. The computing system may generate one or more tag suggestions for the first person portrayed in the image. Each tag suggestion corresponds to a second user from a set of second users associated with the online social network. The one or more tag suggestions may be generated based at least in part on a proximity coefficient indicating a measure of geographical proximity of the first user with respect to each second users, where the proximity coefficient is based on a location history of the first user and a location history of each second user.

What is claimed is: 
     
         1 . A method comprising, by one or more computing systems:
 accessing, by the one or more computing systems, an image associated with an online social network, wherein the image is associated with a first user and portrays at least a first person; and   generating, by the one or more computing systems, one or more tag suggestions for the first person portrayed in the image, each tag suggestion corresponding to a second user from a set of second users associated with the online social network, wherein the one or more tag suggestions are generated based at least in part on a proximity coefficient indicating a measure of geographical proximity of the first user with respect to each second user, wherein the proximity coefficient is based on a location history of the first user and a location history of each second user, wherein:
 the location history of the first user comprises a first set of location updates for a first period of time, each location update in the first set indicating a geographic location of the first user at a particular time within the first period of time; and 
 the location history of the second user comprises a second set of location updates for the first period of time, each location update in the second set indicating a geographic location of the second user at a particular time within the first period of time. 
   
     
     
         2 . The method of  claim 1 , further comprising:
 determining, for each of the set of second users, a facial-recognition score with respect to the first person portrayed in the image, wherein the facial-recognition score is based at least in part on a facial-representation associated with each second user, wherein the facial-representation associated with each second user is compared with the image.   
     
     
         3 . The method of  claim 2 , further comprising determining an affinity coefficient for each second user, wherein the facial-recognition score is further based on the affinity coefficient determined for the first user with respect to each second user. 
     
     
         4 . The method of  claim 3 , wherein generating the one or more tag suggestions for the first person portrayed in the image is further based on the determined facial-recognition scores. 
     
     
         5 . The method of  claim 1 , wherein the proximity coefficient scales with an aggregate value based on, for each location update of the first set of location updates, the distance between a geographic location of the first user and a geographic location of the second user at a concurrent time over the first period of time. 
     
     
         6 . The method of  claim 1 , wherein the proximity coefficient is a weighted combination of multiple subpart proximity coefficients over a period of time, each subpart proximity coefficient being calculated based on the distance between the geographic location of the first user and the geographic location of the second user and a total time that the first user and the second user were at their respective geographical locations. 
     
     
         7 . The method of  claim 1 , wherein one particular second user having the greatest proximity coefficient is suggested for tagging. 
     
     
         8 . The method of  claim 1 , further comprising:
 sending, by the one or more computing systems, to a client system of a third user, instructions for presenting the one or more tag suggestions, each tag suggestion being selectable by the third user to tag the image with the second user corresponding to the tag suggestion.   
     
     
         9 . The method of  claim 8 , wherein the tag suggestions are presented to the third user as a typeahead suggestion. 
     
     
         10 . The method of  claim 9 , wherein the generation of the tag suggestions are updated in response to a character string entered by the third user. 
     
     
         11 . The method of  claim 8 , wherein the first user and the third user are the same user. 
     
     
         12 . The method of  claim 8 , wherein the first user and the third user are different users. 
     
     
         13 . The method of  claim 1 , wherein the image is associated with a particular geographic location. 
     
     
         14 . The method of  claim 13 , wherein a particular second user is eliminated from the tag suggestions if the location history of the particular second user does not correspond to the particular geographic location associated with the image. 
     
     
         15 . The method of  claim 13 , wherein the set of second users is ranked for generating the tag suggestions based on a match between the location history of a particular second user and the particular geographic location associated with the image. 
     
     
         16 . The method of  claim 13 , wherein the particular geographic location associated with the image comprises a location metadata associated with the image on the online social network. 
     
     
         17 . The method of  claim 13 , wherein the particular geographic location associated with the image comprises a location history of a user sharing the image on the online social network, or a location history of a user tagged in the image on the online social network. 
     
     
         18 . The method of  claim 1 , further comprising determining a social affinity for each second user with respect to the first user, wherein generating the one or more tag suggestions is further based on the social affinity determined for each second user with respect to the first user. 
     
     
         19 . One or more computer-readable non-transitory storage media embodying software that is operable when executed to:
 access an image associated with an online social network, wherein the image is associated with a first user and portrays at least a first person; and   generate one or more tag suggestions for the first person portrayed in the image, each tag suggestion corresponding to a second user from a set of second users associated with the online social network, wherein the one or more tag suggestions are generated based at least in part on a proximity coefficient indicating a measure of geographical proximity of the first user with respect to each second user, wherein the proximity coefficient is based on a location history of the first user and a location history of each second user, wherein:
 the location history of the first user comprises a first set of location updates for a first period of time, each location update in the first set indicating a geographic location of the first user at a particular time within the first period of time; and 
 the location history of the second user comprises a second set of location updates for the first period of time, each location update in the second set indicating a geographic location of the second user at a particular time within the first period of time. 
   
     
     
         20 . A system comprising: one or more processors; and a memory coupled to the processors comprising instructions executable by the processors, the processors operable when executing the instructions to:
 access an image associated with an online social network, wherein the image is associated with a first user and portrays at least a first person; and   generate one or more tag suggestions for the first person portrayed in the image, each tag suggestion corresponding to a second user from a set of second users associated with the online social network, wherein the one or more tag suggestions are generated based at least in part on a proximity coefficient indicating a measure of geographical proximity of the first user with respect to each second user, wherein the proximity coefficient is based on a location history of the first user and a location history of each second user, wherein:
 the location history of the first user comprises a first set of location updates for a first period of time, each location update in the first set indicating a geographic location of the first user at a particular time within the first period of time; and 
 the location history of the second user comprises a second set of location updates for the first period of time, each location update in the second set indicating a geographic location of the second user at a particular time within the first period of time.