| CPC G06V 10/443 (2022.01) [G06F 3/0482 (2013.01); G06F 16/9532 (2019.01); G06F 16/9535 (2019.01); G06F 16/9537 (2019.01); G06V 10/751 (2022.01); G06V 20/30 (2022.01)] | 20 Claims |

|
1. A method comprising:
using a client application executing on a user device, generating a user interface configured to display one or more search keywords that are associated with a product that a user is viewing on a screen of the user device;
wherein the one or more search keywords have been associated with one or more corresponding graph of transform invariant features product-tokens (GTIF product-tokens);
displaying the user interface on the user device to depict graphical representations of the one or more search keywords;
determining, by the client application, whether a particular search keyword was selected from the one or more search keywords displayed on the user device;
in response to determining that the particular search keyword was selected from the one or more search keywords displayed on the user device: determining whether an object GTIF product-token, associated with the particular search keyword, matches a particular pair of a set of GTIF product-token pairs;
wherein the set of GTIF product-token pairs comprises one or more of:
a pair comprising a known GTIF product-token and a location data determined for a location of a user device,
a pair comprising known GTIF product-token associated with a user of the user device and one or more social relationships defined for the user,
a pair comprising known time based data associated with one or more events defined for the user and the one or more events,
a pair comprising a known GTIF product-token and a representation of a physical object detected by a camera or sensors and communicated to the user device, or
a pair comprising a known GTIF product-token and a representation of a digital object provided by the user device;
wherein generating the object GTIF product-token comprises: generating an extracted SIFT features by extracting a plurality of local features from an image of the product using a scale-invariant feature transform (SIFT) feature recognition method; generating a SIFT-hash by hashing the extracted SIFT features; and forming the GTIF product-token as a data structure comprising the SIFT-hash;
wherein determining whether the object GTIF product-token matches a particular pair comprises: performing a SIFT-hash comparison by comparing the SIFT-hash of the object GTIF product-token to an index of SIFT-hashes corresponding to known GTIF product-tokens, and identifying a match based on the SIFT-hash comparison;
in response to determining that the object GTIF product-token matches the particular pair, determining particular additional content based on the particular pair, and displaying the particular additional content on the user device.
|