Updating social graph data using physical objects identified from images captured by smartphone

摘要

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.

主权项

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.