Automated content validation and inferential content annotation

Number: 11354894
Published: 2022-06-07
Filed: 2019-10-16
Assignee: Disney Enterprises, Inc.
Inventors: Farre Guiu; Miquel Angel, Petrillo; Matthew C., Alfaro Vendrell; Monica, Fojo; Daniel, Aparicio Isarn; Albert, Guitart Bravo; Francesc Josep, Badia Pujol; Jordi, Martin; Marc Junyent, Accardo; Anthony M.
CPC: G06N20/00; G06V10/774; G06N5/046; G06F18/214; G06V10/7784; G06V20/70; G06N5/02; G06V20/41; G06N5/025; G06V20/20; G06N5/04
Verdict: Set aside content validation/annotation, business
Source: Google Patents · FreePatentsOnline

Abstract

According to one implementation, a system for automating inferential content annotation includes a computing platform having a hardware processor and a system memory storing a software code including a set of rules trained to annotate content inferentially. The hardware processor executes the software code to utilize one or more feature analyzer(s) to apply labels to features detected in the content, access one or more knowledge base(s) to validate at least one of the applied labels, and to obtain, from the knowledge base(s), descriptive data linked to the validated label(s). The software code then infers, using the set of rules, one or more label(s) for the content based on the validated label(s) and the descriptive data, and outputs tags for annotating the content, where the tags include the validated label(s) and the inferred label(s).

Background

BACKGROUND (1) Audio-visual media, such as video, continues to be a highly popular media format for the enjoyment of entertainment content in the form of movie, television, sports, and gaming content, for example. Due to its popularity with consumers, ever more audio-visual content is being produced and made available for distribution. Consequently, the accuracy and efficiency with which such content can be reviewed, classified, archived, and managed has become increasingly important to its producers, owners, and distributors. Consequently, there is an ongoing need in the art for automated solutions for annotating or “tagging” audio-visual content in order to reduce the time and costs expended on the production and management of that content. SUMMARY (2) There are provided systems and methods for automating content validation and inferential content annotation, substantially as shown in and/or described in connection with at least one of the figures, and as set forth more completely in the claims.

Claims

1. An automated content annotation system comprising: a computing platform including a hardware processor and a system memory; a software code stored in the system memory, the software code including a set of rules trained to annotate content inferentially; the hardware processor configured to execute the software code to: utilize at least one feature analyzer to apply a plurality of labels to features detected in the content; access at least one knowledge base to validate at least one of the applied plurality of labels; obtain, from the at least one knowledge base, a descriptive data linked to the at least one validated label in the at least one knowledge base; infer, using the set of rules, at least one label for the content based on the at least one validated label and the descriptive data linked to the at least one validated label; and output tags for annotating the content, the tags including the at least one validated label and the inferred at least one label. || 11. A method for use by an automated content annotation system including a computing platform having a hardware processor and a system memory storing a software code including a set of rules for trained to annotate content inferentially, the method comprising: utilizing at least one feature analyzer, by the software code executed by the hardware processor, to apply a plurality of labels to features detected in the content; accessing at least one knowledge base, by the software code executed by the hardware processor, to validate at least one of the applied plurality of labels; obtaining, by the software code executed by the hardware processor, from the at least one knowledge base, a descriptive data linked to the at least one validated label in the at least one knowledge base; inferring, by the software code executed by the hardware processor and using the set of rules, at least one label for the content based on the at least one validated label and the descriptive data linked to the at least one validated label; and outputting, by the software code executed by the hardware processor, tags for annotating the content, the tags including the at least one validated label and the inferred at least one label.