Systems and methods for automatically generating headshots from a plurality of still images
The present disclosure relates to systems and methods for generating headshots from a plurality of still images. In one implementation, the system may include a memory storing instructions and a processor configured to execute the instructions. The instructions may include instructions to receive a plurality of still images from one or more video feeds, score the plurality of images along a plurality of dimensions based on a scale, rank the plurality of images using at least one of a composite score or at least one of the dimensions, select a subset of the plurality of images using the ranking, and construct at least one headshot of the user from the subset of the plurality of images.
1. A system for updating a user headshot, the system comprising:
a memory storing instructions; and
a processor configured to execute the instructions to:
receive, from a social network, feedback associated with a headshot,
adjust a classification of the headshot based on the feedback,
modify an image classifier associated with the classification based on the feedback,
classify, by the modified image classifier, a set of headshots of a user,
select a headshot from the set of headshots based on the classifications of the set of headshots and a context associated with the user, and
upload, to a server, the selected headshot as an updated headshot of the user.
2. The system of claim 1 , wherein the instructions further comprise instructions to map the received feedback to an identifier of the headshot using a database of headshots.
3. The system of claim 2 , wherein the instructions further comprise instructions to remove the headshot from the database in response to the feedback.
4. The system of claim 2 , wherein the instructions further comprise instructions to transmit a new headshot to the social network in response to the feedback.
5. The system of claim 1 , wherein the adjustment of the classification is further based on an identity of the social network.
6. The system of claim 1 , wherein the instructions further comprise instructions to:
aggregate a plurality of feedback associated with the headshot from the social network,
wherein the adjustment of the classification and the modification of the image classifier is further based on the aggregated feedback.
7. The system of claim 6 , wherein the aggregation is performed over a predetermined period of time.
8. The system of claim 1 , wherein the instructions further comprise instructions to:
aggregate a plurality of feedback associated with the headshot from a plurality of social networks,
wherein the adjustment of the classification and the modification of the image classifier is further based on the aggregated feedback.
9. The system of claim 8 , wherein the aggregation is performed over a predetermined number of social networks.
10. The system of claim 1 , wherein the feedback originates from a user that is not depicted in the headshot.
11. A computer-implemented method for updating a user headshot, the method comprising:
receiving, from a social network, feedback associated with a headshot,
adjusting a classification of the headshot based on the feedback, and
modifying an image classifier associated with the classification based on the feedback,
classifying, by the modified image classifier, a set of headshots of a user,
selecting a headshot from the set of headshots based on the classifications of the set of headshots and a context associated with the user, and
uploading, to a server, the selected headshot as an updated headshot of the user.
12. The method of claim 11 , wherein the feedback comprises a reaction posted on the social network by another user in response to the headshot.
13. The method of claim 11 , wherein the feedback comprises text posted on the social network by another user in response to the headshot.
14. The method of claim 13 , further comprising:
processing the feedback using natural language processing (NLP),
wherein the adjustment of the classification and the modification of the image classifier is based on an output of the NLP.
15. The method of claim 14 , wherein the feedback comprises at least one of data regarding how many users viewed the headshot or lengths of time for which users viewed the headshot.
16. A non-transitory computer-readable medium storing instructions that, when executed by one or more processors, cause the one or more processors to:
receive, from a social network, feedback associated with a headshot,
adjust a classification of the headshot based on the feedback, and
modify an image classifier associated with the classification based on the feedback,
classify, by the modified image classifier, a set of headshots of a user,
select a headshot from the set of headshots based on the classifications of the set of headshots and a context associated with the user, and
upload, to a server, the selected headshot as an updated headshot of the user.
17. The non-transitory computer-readable medium of claim 16 , wherein modifying the image classifier comprises adjusting the image classifier to reduce an associated loss function.
18. The non-transitory computer-readable medium of claim 17 , wherein the loss function comprises at least one of a square loss function, a hinge loss function, a logistic loss function, a cross entropy loss function, or a combination thereof.
19. The non-transitory computer-readable medium of claim 18 , wherein the loss function is selected based on one or more properties of the image classifier.
20. The non-transitory computer-readable medium of claim 18 , wherein the loss function is selected by one or more models trained to select loss functions.