Updating social graph data using physical objects identified from images captured by smartphone
First Claim
Patent Images
1. A method comprising the acts:
- analyzing first data corresponding to imagery captured by a user'"'"'s smartphone camera, to yield identification data corresponding to a first non-human physical object depicted in the imagery, said analyzing being performed using electronic circuitry and comprising an image fingerprinting operation that includes (a) identifying candidate keypoints by convolving the first data with a blur filter to detect space-scale extrema of a scale-localized Laplacian transform of the image, (b) discerning and discarding keypoints having a contrast below a threshold value, said discerning including performing a second order Taylor expansion, and (c) checking remaining keypoints for matches with reference keypoints for reference physical objects; and
issuing instructions that a social network graph data structure be updated to add node and tie elements, the node element corresponding to the first physical object, and the tie element indicating interaction by the user with said first physical object, information in said social network graph data structure being published by a service for viewing by one or more third parties;
wherein the first physical object is identified by reference to the first data, and the user'"'"'s interaction with the first physical object is automatically publicized to said one or more third parties.
0 Assignments
0 Petitions
Accused Products
Abstract
This technology concerns, in one aspect, using a person'"'"'s social network graph data as a virtual visual cortex—taking image input from a smartphone or the like, and processing it with the graph data to yield a personalized form of processing based on the imagery. The user'"'"'s network graph data is typically updated by such processing—providing a form of virtual image memory that can influence future social network behavior. In another aspect, the technology concerns identifying content (e.g., audio) by both fingerprint-based and watermark-based techniques, and arrangements employing such identification data. A great number of other features and arrangements are also detailed.
64 Citations
21 Claims
-
1. A method comprising the acts:
-
analyzing first data corresponding to imagery captured by a user'"'"'s smartphone camera, to yield identification data corresponding to a first non-human physical object depicted in the imagery, said analyzing being performed using electronic circuitry and comprising an image fingerprinting operation that includes (a) identifying candidate keypoints by convolving the first data with a blur filter to detect space-scale extrema of a scale-localized Laplacian transform of the image, (b) discerning and discarding keypoints having a contrast below a threshold value, said discerning including performing a second order Taylor expansion, and (c) checking remaining keypoints for matches with reference keypoints for reference physical objects; and issuing instructions that a social network graph data structure be updated to add node and tie elements, the node element corresponding to the first physical object, and the tie element indicating interaction by the user with said first physical object, information in said social network graph data structure being published by a service for viewing by one or more third parties; wherein the first physical object is identified by reference to the first data, and the user'"'"'s interaction with the first physical object is automatically publicized to said one or more third parties. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17)
-
-
18. A method comprising the acts:
-
analyzing first data corresponding to imagery captured by a user'"'"'s smartphone camera, to yield identification data corresponding to a first non-human physical object depicted in the imagery, said analyzing comprising an image fingerprinting operation that includes (a) identifying candidate keypoints by convolving the first data with a blur filter to detect space-scale extrema of a scale-localized Laplacian transform of the image, (b) discerning and discarding keypoints having a contrast below a threshold value, said discerning including performing a second order Taylor expansion, and (c) checking remaining keypoints for matches with reference keypoints for reference physical objects; issuing instructions that a social network graph data structure be updated to add node and tie elements, the node element corresponding to the first physical object, and the tie element indicating interaction by the user with said first physical object; analyzing second data corresponding to imagery captured by the user'"'"'s smartphone camera to yield identification data corresponding to a second physical object depicted in said imagery, said analyzing comprising an image fingerprinting operation that includes (a) identifying candidate keypoints by convolving the second data with a blur filter to detect space-scale extrema of a scale-localized Laplacian transform of the image, (b) discerning and discarding keypoints having a contrast below a threshold value, said discerning including performing a second order Taylor expansion, and (c) checking remaining keypoints for matches with reference keypoints for reference physical objects; issuing instructions that a social network graph data structure be updated to add node and tie elements, the node element corresponding to the second physical object, and the tie element indicating interaction by the user with said second physical object; and storing rule data, the rule data establishing that the user'"'"'s interaction with the first physical object may be published to a first group of third parties, and that the user'"'"'s interaction with the second physical object may not be published to at least certain parties of said first group; wherein said analyzing acts are performed using electronic circuitry.
-
-
19. A method comprising the acts:
-
using first software and electronic circuitry, analyzing first data corresponding to first content captured by a sensor in a user'"'"'s smartphone from a first non-human object, to yield first identification data for the first object, said analyzing comprising an image fingerprinting operation that includes (a) identifying candidate keypoints by convolving the first data with a blur filter to detect space-scale extrema of a scale-localized Laplacian transform of the image, (b) discerning and discarding keypoints having a contrast below a threshold value, said discerning including performing a second order Taylor expansion, and (c) checking remaining keypoints for matches with reference keypoints for reference physical objects; issuing instructions that a social network graph data structure be updated to add node and tie elements, the node element corresponding to the first object, and the tie element indicating interaction of the user with said first object; storing at least some first content data in association with said node; using second software, analyzing said first content data to obtain second identification data, the second identification data being different than the first identification data; and issuing instructions that said social network graph data structure be updated to reflect said second identification data. - View Dependent Claims (20, 21)
-
Specification