New pending patent applications are published on Thursdays. This week's Spotlight Applications concern aspects of digital fingerprinting. Assigned to EMC, the first application discloses techniques for fingerprint based status detection in a distributed processing system. Assigned to Rovi, the second application discloses techniques for identifying a source of media content based on a log of fingerprints.
20170093672, "Method and device for fingerprint based status detection in a distributed processing system," assigned to EMC.
A method, device, and computer program product for fingerprint based status detection in a distributed processing system is provided. The method comprises: generating and sending, at a root node, an initial fingerprint based on the output message ID to be sent; at a middle node, generating and sending an updated fingerprint based on a received input message ID, an output message ID to be sent and a received fingerprint; sending, at the leaf node, a final updated fingerprint to a tracking task point; and obtaining, at the tracking task point, status information based on the final updated fingerprint. As the method does not generate traffic at each node but only generates tracking traffic at leaf nodes, the approach according to the present disclosure reduces tracking overhead significantly while guaranteeing the reliability of data processing.
20170094347, "Systems and methods for identifying a source of media content based on a log of fingerprints," assigned to Rovi.
Systems and methods are described for identifying a content source of media content being transmitted to a viewer in situations when a specific media asset is being transmitted to the viewer from two different content sources nearly simultaneously. For example, if a certain movie is being broadcast at nearly the same time by two different channels, the system is able to identify the channel that the viewer is watching. The system performs the identification by obtaining fingerprints for two different channels, matching the fingerprints based on time and then finding a divergent pair of fingerprints by iterating in reverse chronological order through the pairs and comparing the fingerprints. The system then compares the fingerprints to fingerprints in a log that stores fingerprints and corresponding generation times for content that the viewer is consuming. The matching fingerprint corresponds to the correct content source.