The patent badge is an abbreviated version of the USPTO patent document. The patent badge does contain a link to the full patent document.
The patent badge is an abbreviated version of the USPTO patent document. The patent badge covers the following: Patent number, Date patent was issued, Date patent was filed, Title of the patent, Applicant, Inventor, Assignee, Attorney firm, Primary examiner, Assistant examiner, CPCs, and Abstract. The patent badge does contain a link to the full patent document (in Adobe Acrobat format, aka pdf). To download or print any patent click here.
Patent No.:
Date of Patent:
Aug. 23, 2016
Filed:
Sep. 08, 2011
Joakim Söderberg, Solna, SE;
Che Zhong, Solna, SE;
Joakim Söderberg, Solna, SE;
Che Zhong, Solna, SE;
Telefonaktiebolaget LM Ericsson (Publ), Stockholm, SE;
Abstract
There is described a device and method for assigning a tag to a media file associated with metadata describing features of the media file. The metadata is converted into a query feature vector associated with the media file. A weighted distance is calculated between the query vector and each of a plurality of tag feature vectors, where each tag feature vector is included in a tag instance stored in a database, each tag instance further comprising a tag and a weight vector. The weighted distance is determined by comparing each feature of the query vector with the corresponding feature in the tag feature vector to obtain an initial distance value for that feature, modifying the initial distance value by a weight for that feature obtained from the weight vector in the tag instance to obtain a comparison value for that feature, and combining the comparison values of all of the features to obtain the weighted distance. Those tag instances whose tag feature vectors have the smallest weighted distance from the query vector are selected. The user chooses one or more of the selected tags, which are then associated with the media file. The weight for each feature in the weight vectors in the tag instances selected by the tag selection unit is updated on the basis of whether or not the tag in each tag instance was chosen by the user, and the distance between the feature of the tag feature vector and the corresponding feature in the query vector.