none
grey

Machine learning based video segmentation

With the broadcast industry making a transition from TV to OTT, spinning up new linear channels or making content available to VOD platforms at faster pace has become imperative. TV networks and OTT content creators today struggle with twin problem of growing the user-base and maintaining active subscribers to be economically viable.

Key content challenges for broadcasters in 2017 and beyond

  • Scaling up business without incurring upfront cost
  • Implementing a factory-scale model for processing long form content
  • Reduce manual errors with automation
  • Create innovative business models to monetize new content formats

Limitations of traditional channel playouts and content segmentation services

Traditional broadcast infrastructure and managed services tend to be asset-heavy, and manual in nature. The traditional broadcast workflow involves integrating with third party tool-chain, and using on-premise stores that restrict the scalability and turnaround time for broadcasters. Additionally, processing large multimedia files for broadcast could involve deploying large teams for quality checks, and segmentation. In financial terms, this could be the biggest roadblock that TV channels and content creators face when it comes to expanding faster across geography and formats

Media Management and segmentation simplified with Machine Learning & Automation

Processing large volumes of media files requires a combinatorial approach. Amagi’s TORNADO content preparation service can ensure that TV networks can scale-up faster and make their content available on multiple platforms with minimum manual intervention.

Benefits of using Machine Learning Services

  • Reduced turnaround time by automation driven quality control
  • Automated content segmentation, ready for Future TV
  • Improved cost management with reduced human intervention

Amagi’s automated, cloud-based content & metadata aggregation

Amagi’s Machine Learning System can automatically aggregate content and metadata from various distributors all over the world. The system can dynamically convert formats using a format translator.

Once the content is acquired, Amagi’s system can perform automated QC to ensure consistent quality for content and metadata. Based on the requirement, it can perform specific actions such as adding or removing color bars, countdown clock/slate, or create segments as per instructions. Finally, the system uses a transcode farm to deliver the aggregated content and metadata in desired format, while also enabling secure access with a user-friendly web UI.

Broadcast Quality Control Services Offered by Amagi

  • Syntactic QC Examples
    • Black frame of frozen frame detection
    • Missing Audio/Silence detection
    • Verification of audio video sync at regulated intervals
    • Content and metadata format errors
    • Subtitle timecode verification
    • Loudness error (LKFS verification)
  • Semantic deep-learning based quality check examples
    • Check if subtitles are contextually in sync with AV content
    • Check if summary of content is in sync with AV content
  • Manual QC
    • Manual verification of results from the automated QC to ensure greater accuracy

Machine Learning based content segmentation services

  • Automatically creates content segments from long-form video, based on specifications provided by the customer
  • Create promos from the video for marketing purpose
  • Suggest logical ad-break points or segments
  • Extract metadata from the video automatically

Getting started

Talk to Amagi’s broadcast playout consultants, to discuss Machine Learning based content segmentation services for your TV channel. Drop us a note!

Interested in:






Back to Top