- Number
- 12279003
- Published
- 2025-04-15
- Filed
- 2022-11-18
- Assignee
- Disney Enterprises, Inc.
- Inventors
- Strein; Michael J.
- CPC
- H04N21/2405; H04N21/2543; H04N21/64769
- Verdict
- Set aside cloud production infrastructure, plumbing
- Source
- Google Patents · FreePatentsOnline
Abstract
A system includes a processor, and a memory storing software code and a machine learning (ML) model trained to allocate media production resources. The processor executes the software code to receive data describing a media flow requiring processing, identify, using the data and the ML model, media production resources for processing the described media flow, obtain the media production resources, and aggregate, from each of the media production resources, performance and billing metrics of a respective one of the media production resources resulting from processing of the described media flow by the media production resources. The processor may further execute the software code to determine, using the aggregated performance and billing metrics, a resource allocation efficiency score corresponding to each of one or more of the media production resources to provide one or more resource allocation efficiency score(s), and further train, using the resource allocation efficiency score(s), the ML model.
Background
BACKGROUND (1) Media production facilities typically rely on local software defined networks (SDNs) to efficiently move media flows and coordinate media production. In traditional media production environments, these SDNs for managing media flows are implemented using “on-premises” resources for which the production facility has control of the information technology (IT) architecture. In addition, the actual work of media production is often performed using on-premises hardware resources and on-premises human expertise. (2) However, as highly scalable cloud-based virtual resources become increasingly available and affordable, and as automation solutions become increasingly proficient in replicating the performance of human contributors to media production, the transition to cloud-based and hybrid cloud/on-premises (hereinafter “cloud and hybrid-cloud”) media production grows more attractive. However, in the present cloud-based resource environment, production resources are typically provided a la carte by a variety of different vendors, so that a media production facility must normally obtain computing and storage resources under a contract with a particular vendor, arrange for media flow transfer bandwidth from another vendor, license media production software applications from yet other vendors, and so forth. Consequently, there is a need in the art for systems providing consolidated cloud and hybrid-cloud production management.
Claims
1. A system comprising: a processor; and a memory storing a software code and a machine learning (ML) model trained to allocate a plurality of media production resources; the processor configured to execute the software code to: receive data describing a multicast media flow including audio-video content and originating from one or more senders to be provided to a group of receivers, the data including a size of the multicast media flow, a type of content of the multicast media flow, and a time constraint; identify, using the data and the ML model, the plurality of media production resources for use in processing the audio-video content; obtain the plurality of media production resources for processing the audio-video content; process the audio-video content by the plurality of media production resources; after processing, provide the processed audio-video content from the one or more senders to the group of receivers; and aggregate, from each of the plurality of media production resources, performance and billing metrics of a respective one of the plurality of media production resources resulting from processing of the audio-video content by the plurality of media production resources. ||
8. A method for use by a system including a processor and a memory storing a software code and a machine learning (ML) model trained to allocate a plurality of media production resources, the method comprising: receiving, by the software code executed by the processor, data describing a multicast media flow including audio-video content and originating from one or more senders to be provided to a group of receivers, the data including a size of the multicast media flow, a type of content of the multicast media flow, and a time constraint; identifying, by the software code executed by the processor and using the data and the ML model, the plurality of media production resources for use in processing the audio-video content; obtaining, by the software code executed by the processor, the plurality of media production resources for processing the audio-video content; processing, by the software code executed by the processor, the audio-video content by the plurality of media production resources; after processing, providing, by the software code executed by the processor, the processed audio-video content from the one or more senders to the group of receivers; and aggregating, from each of the plurality of media production resources, by the software code executed by the processor, performance and billing metrics of a respective one of the plurality of media production resources resulting from processing of the audio-video content by the plurality of media production resources. ||
15. A system comprising: a processor; and a memory storing a software code and a machine learning (ML) model trained to allocate a plurality of media production resources; the processor configured to execute the software code to: receive data describing a media flow requiring processing, the data including a type of content of the media flow; identify, using the data and the ML model, the plurality of media production resources for use in processing the described media flow; obtain the plurality of media production resources; aggregate, from each of the plurality of media production resources, performance and billing metrics of a respective one of the plurality of media production resources resulting from processing of the described media flow by the plurality of media production resources; determine, using the aggregated performance and billing metrics, a resource allocation efficiency score corresponding to each of one or more of the plurality of media production resources to provide one or more resource allocation efficiency scores; and further train, using the one or more resource allocation efficiency scores, the ML model trained to allocate the plurality of media production resources.