Articles

Thought leadership articles by IABM and our members
Articles taken from IABM's journal and at show papers
To submit your article email marketing@theiabm.org

BaM™ Product Highlight: Amagi

Mon 09, 07 2018

MANAGE CATEGORY BaM™Award Nominee – Amagi – Heralding an era of cognitive playout infrastructure with machine learning-augmented content preparation services

Amagi - Heralding an era of cognitive playout infrastructure with machine learning-augmented content preparation services

IABM BaM Awards

Over the last three years, broadcast industry trends are signaling a few important directives for TV networks.

  1. The rise of individualism is creating an unprecedented demand for multi-screen content consumption
  2. The millennial generation’s attitude of ‘here and now’ is pushing the innovation boundaries in content discovery and delivery
  3. The changing audience loyalty is further impacting subscription and advertising revenues

In such an evolving scenario, TV networks, content owners, and digital first networks are creatively trying to grab the piece of the action by building their presence across multiple content delivery platforms. For example, traditional TV networks are now adding linear OTT streams to their portfolio. Whereas, digital first networks are vying against traditional networks by delivering their feeds to more traditional cable distribution platforms. Also, content owners are leveraging their vast libraries to set up their own linear channels and delivering to VOD platforms. While the midstream activities of playout, and downstream activities of delivery have witnessed significant innovation with the emergence of cloud as a reliable broadcast technology option, upstream workflow activities of content preparation have remained manually intensive.

IABM BaM Awards Product Highlight - Amagi

Traditional content preparation modules involve large physical infrastructure to store content and accommodate armies of video specialists to review content assets. This process is not only expensive, but is very time-consuming. As most of the work in this phase is routine and repeatable, it becomes an ideal candidate for automation saving precious man-hours, and more importantly scale rapidly.

Amagi TORNADO is a first-of-its-kind, cloud-based machine learning-augmented content preparation service that addresses requirements of efficiency and scale with equal ease. It is game-changing because of its applicability to a wide variety of content preparation activities catering to the unique needs of TV networks, content owners, vMVPD platforms, and digital first networks. Amagi TORNADO is conceptualized as a family of machine learning-based content preparation services whose scope can be expanded with increase in machines learning more about each segment of a video asset as they process higher volumes of content.

At the moment, Amagi TORNADO focuses on three high impact areas that can deliver both top-line revenue growth, and bottom-line savings.

Factory-scaling of VOD segment creation

Linear broadcast model is highly reliant on sophisticated processing of video for ad break points identification, credits, color bars and blacks. Using Amagi TORNADO, video assets can be uploaded on to an Amagi-supported public cloud infrastructure such as AWS and Microsoft Azure. Amagi TORNADO synthesizes each segment in a video, intelligently learns from the inputs of human content preparation specialists and prepare segments for playout with rapid turnaround time. This is accomplished by training the platform to identify logical break-points through understanding of scene change and audio levels. The entire workflow of uploading content assets, scheduling assets for preparation, segmenting, and status reporting can be controlled from any remote location using a web-based UI.

Amagi TORNADO synthesizes each segment in a video, intelligently learns from the inputs of human content preparation specialists and prepare segments for playout with rapid turnaround time - MANAGE CATEGORY BaM™Award Nominee: Amagi Click To Tweet

While taking a traditional approach to VOD segmentation, the rule of thumb allocates one man-hour effort to segment one hour of content. With Amagi TORNADO, this time is drastically reduced to less than 10 minutes per hour of content. Amagi TORNADO targets a 1:6 efficiency ratio vis-a-viz 1:1 ratio of manual VOD segmentation. As a result, by using Amagi TORNADO, TV networks can spin up large number of servers on the cloud to process high volumes of content simultaneously, increasing speed while reducing cost of VOD segmentation.

Near real-time live to VOD conversion

One of the use cases that Amagi TORNADO can potentially address is the near real-time creation of VOD content from live broadcast. For example, in case of sports broadcast, using machine learning Amagi TORNADO can create a highlights package from a fresh off-the-air sports match, capturing key moments of the game – start, goals, player injury, referee decisions, final whistle and so on, and deliver it to OTT platforms in just minutes to enable immediate post-match binging through VOD. It can also reduce subscriber churn dramatically and lend scale to VOD creation from 100s of concurrent live channels, without deploying manual support.

Auto ad detection of mid-roll ads

OTT today presents the biggest revenue generation opportunity in term of mid-roll ad insertion. In case of vMVPD platforms who are aggregating feeds for further distribution, very few broadcasters provide them feeds with ad markers. As a result, vMVPD platforms are unable to identify mid-roll ads and replace them locally. This leads to loss of additional ad revenues. Using machine learning techniques, the Amagi service can automatically detect ads in linear broadcast streams by comparing video segments with an active ad library. With repeated exposure to volumes of ads, the accuracy and speed of detection can attain efficiency levels in excess of 80 percent. The service can be integrated with Amagi’s server-side ad insertion platform to dynamically replace detected ads with targeted ads and deliver it as a unified stream. Using such machine learning techniques, vMVPDs can now build a scalable ad revenue model by preparing for mid-roll ad insertions in input content streams.

Using machine learning techniques, the Amagi service can automatically detect ads in linear broadcast streams by comparing video segments with an active ad library - MANAGE CATEGORY BaM™Award Nominee: Amagi Click To Tweet

By deploying Amagi TORNADO, TV networks and content creators can effectively complement the growth of cognitive playout infrastructure, delivering a truly innovative, factory-scale model of content preparation for broadcast.

Search For More Content



X