OSN - International Regional Compliance

osn-logo_transparent.png

OSN is a DTH satellite & OTT provider in the Middle East. They receive content from around the world with different compliance standards and manually check each asset. Our client was looking for a way to accelerate compliance with applicable laws, efficiently flag potential issues, objectively defend decisions to regulators, and identify issues for deciding ratings.

SOLUTION

GrayMeta Platform enables OSN to quickly find key moments in its content and automatically generate metadata associated with a range of use cases by using Machine Learning and Artificial Intelligence.  Our solution not only highlighted Ofcom related compliance issues but our custom face model and brand detection allowed our client to find instances where local dignitaries or local logos appeared in the timeline – essential for regional compliance.

Curio UI.png

VALUE

  • Reduce human error and increase operational efficiency

  • Automate current and new workflows around compliance

  • Achieve resiliency and scale with a modern cloud infrastructure

  • Reduce legal exposure due to inappropriate or contentious content being aired on our platform

Below Dave Mace from OSN discusses how they approached the adoption and leveraging ML and AI in their media supply chain

 
We reduced the amount of footage our human curators needed to review by over a third.
— Dave Mace, Digital and Cloud Services, OSN
donald-giannatti-Wj1D-qiOseE-unsplash.jpg
 
 

VIDEOFASHION - ACCESS & MONETIZING 40 YEARS OF FASHION

video fashion.png


videofashion-still.png


VideoFashion, the world’s premiere fashion video producer and supplier had a large content archive with over 18,000 hours of historical footage.

They needed to enrich their content and improve their content discovery, allowing them to ultimately improve content production and editorial processes to leverage new monetization opportunities.

But first they needed to get their archive off tape and onto file.

SOLUTION

Before VideoFashion’s archive could be properly monetized, we helped create a strategy for tape migration to file and into the cloud.

vf storage boxes 01.png
vf storage boxes 02.png

By using various machine learning services (such as facial recognition, logo/brand detection, OCR, Speech to Text and GrayMeta's Technical Services), they were able to gain a deep understanding of their content, and quickly surface clips they would otherwise be unable to find.

VF Tags UI.png

By combining different machine learning metadata, VideoFashion is now able to search through their archive using GrayMeta Curio and find suitable clips in seconds.

VALUE

  • Reduction of manual workflows

  • Improved customer experience

  • Increase in monetization opportunities and speed to sales of existing content

  • Improved search and discovery, through expansive metadata

Learn more about how VideoFashion has unlocked the value of their 40 year collection of Fashion content by watching their case study video below

 
 
I am constantly astounded by how quickly I can find what I need using Curio.
— Anne Adami, President VideoFashion
Fendi+-+Colored.jpg
 
 

Rock & Roll Hall of Fame - Archive, Search, Discovery and Access

twitter_card.png

ROCK & ROLL HALL OF FAME

The Rock & Roll Hall of Fame recognizes and archives the history of the best-known and most influential music artists, producers, engineers, and other notable figures who have had a major influence on the development of rock and roll.  They have decades of content stored in multiple locations and were unable to quickly locate and fully utilize these valuable assets. 

SOLUTION

GrayMeta created a single, searchable index of data from all of Rock Hall’s archives. Using GrayMeta’s Curio Enterprise, the technical and machine learning metadata allowed the team to understand their assets and quickly search through images, audio, and video. With a rich archive of artists and figures, facial training and recognition was critical for Rock Hall’s use cases. This metadata is being integrated into their MAM solution.

VALUE

  • Reduced time searching for relevant clips

  • Increased ability to fully exploit their rich and varied archive

  • Improved operational efficiency by creating automated tags for content

 
 
 
“Using Curio is a more efficient way to discover and retrieve iconic music moments from our rich collection of music history in a matter of minutes. We are re-discovering valuable assets like needles in an organized rock and roll haystack.”
— Heidi Quicksilver Rock & Roll Hall of Fame
 
applause-audience-concert-2147029.jpg
 
 

SKY UK - QC OPERATIONS

sky-plc-vector-logo-small.png




Sky UK needed to conduct technical quality checks on an increasing volume of UHD, HDR and IMF packaged content.

SOLUTION

Sky UK uses the GrayMeta Iris Media Platform for all their UHD, HDR and IMF content workflows. The GrayMeta Iris Media Platform is an enterprise media playback and QC solution with open API’s which enables broadcasters, studios and content owners to quickly and accurately collaborate and deliver content through the content life cycle which drives time and cost efficiencies for customers.

Iris-QC-Pro_DHQC.png


Iris delivers true reference playback of complex IMF packages and other computationally heavy files in both SDR and HDR formats.

VALUE

  • Confidence in UHD, HDR and IMF accuracy.

  • Improve automation and remove the need for a QC operator to manually comb through lines of technical data for comparison of in-house or external requirements. 

  • Improved collaboration across the Sky operation QC teams, removing inefficiencies and costs while reducing time to QC.

 

 
 
GrayMeta has helped us handle the complexities of IMF with ease and validate important HDR metadata. The reference player is used to drive quality control through our supply chain and provides us the confidence that our viewers are getting the best experience when they watch our UHD content
— Ashley Ross, Technical Operations Manager, Sky
annie-spratt-g9KFpAfQ5bc-unsplash.jpg
 
 
 
 

Sky News - Royal Wedding

6523.1485317926.png




MARKETING AND CUSTOMER ENGAGEMENT : Royal wedding, Sky UK

Sky News wanted to leverage their coverage of the Royal Wedding to create a differentiated storytelling experience for their viewers and maximize engagement on their digital platforms.

Solution

With the Curio Live Solution and the Platform UI, the network was able to power its “Royal Wedding: Who’s Who Live” app with accurate and live facial recognition results.

Sky News Royal Wedding Who’s Who

Sky News Royal Wedding Who’s Who

GrayMeta, together with AWS and UI Centric designed and developed a live and video on demand solution. GrayMeta developed a custom AI face model, sourcing and curating images from an anticipated guest list. To ensure 100% accuracy, Curio Live provided a live human curation tool for the event.

value

  • Award winning digital storytelling experience

  • Increased global media impressions to the Sky News website

  • Combined innovative human curation of AI with live analysis to ensure accuracy


Sky continuously searches for ways to innovate and bring better coverage to its customers. This new functionality allows Royal Wedding viewers greater insight into one of the biggest live events of the year, wherever they are. We’re excited by the software’s potential and ability to give audiences new ways of consuming content.
— David Gibbs, Sky News
 

king-s-church-international-vsk-Vw8AXe4-unsplash.jpg
 
 

International post-production house: Accelerating Offline Editing

noun_time management_1327062.png

74% of creatives say that more than half of their time is spent on repetitive non-creative tasks

An international post-production house based in London that produces short-form content for brands and advertisers.

Our client wanted to reduce production costs by automatically tagging footage captured at location shoots while providing the ability to search across historic shoots and finished commercials to find opportunities to re-purpose content.

SOLUTION 

Dailies from the location shoot were uploaded to secure cloud storage so that an off-line editor could access them for post-production. By using a combination of machine learning services (such as speech-to-text, visual tags, facial recognition, and logo detection) GrayMeta Curio was used to automatically tag dailies.

 
curio-known-graphics-shadow.png

Known Graphics & Logo Detection

Identify brands within the video using GrayMeta Curio

 
curio-people-shadow.png

Face Detection & Recognition

Train custom faces or recognize celebrities using GrayMeta Curio

 
curio-speech-to-text-shadow.png

Speech to Text

Generate a transcript based on speech to text using GrayMeta Curio

 
curio-tags-descriptions-shadow.png

Visual Tags & Descriptions

Recognize visual objects and scenes using GrayMeta Curio

Remote off-line editors used the GrayMeta Curio extension for Adobe® Premiere® Pro to view this automatically generated metadata in a timeline to assemble final adverts.

The Curio extension for Adobe Creative Cloud enables Curio users to simply drag-and-drop clips and images into Adobe Premiere Pro, Adobe Photoshop, Adobe Illustrator and Adobe InDesign – saving content creators valuable time.

See the video below for more information about the Curio extension for Adobe Creative Cloud:

 VALUE

  • Improved productivity at shoots

  • Reduce offline edit turnaround time and time searching for clips

  • Maximize the investment made at every shoot

  • Speed up localization of content for different markets

 
 
Our offline editing process is faster, our international offices can collaborate better, and we have a way of quickly finding assets quickly allowing us to re-purpose content.
— VFX Technical Supervisor, International Post-Production
Adobe.png
jc-gellidon-3uqauucYhjQ-unsplash.jpg
 
 

Public Broadcaster Using AI to Drive Contextual Advertising

noun_Productivity_252395 (1).png

AI technologies are projected to increase labor productivity by up to 40%

public broadcaster Contextual ADvertising

A UK Public Broadcaster conducted research with a prominent UK university that showed consumer brand retention was improved significantly when adverts were placed contextually. Our client wanted to provide contextual ad placements for its brands on its linear channels as part of a pilot. Using the GrayMeta Platform, contextual and time specific metadata was created to describe scenes shortly before a linear ad break and our client used this information to power the placement of linear adverts.

SOLUTION 

GrayMeta created time-specific content metadata using a range of machine learning services to identify key moments and objects in their assets. This metadata was integrated into the broadcaster’s demand side sales platform, allowing for contextually relevant ad placements, driving ad revenue and improving consumer experience.

VALUE

  • Operational efficiency through automatic tagging of content

  • Improved brand retention

  • Better consumer experience with contextually placed adverts

 
 
jeshoots-com-PpYOQgsZDM4-unsplash.jpg
 
 

Finding Clips in Sports Archives for Social Media

using machine learning to generate metadata for sports archive content

Mining recent footage and archives for clips that can engage fans on social media is increasingly important for sports federations and clubs, whether football or esports. Finding relevant video, audio and images as quickly as possible is an essential part of a social media strategy.

Lack of metadata and content stored across multiple storage locations often means that organizations cannot fully exploit their sports archives. The length of time is takes to find relevant assets means that its too late to post or loses its value.

SOLUTION 

The GrayMeta Platform connects to multiple storage locations within a sports archive and analyses content. GrayMeta creates time-specific metadata relevant to sports content using a range of machine learning services, including speech to text, OCR, and face recognition. GrayMeta is able to combine all these machine learning services with its own custom soccer model to detect key events in a match.

soccer model 1.png

Custom Soccer Model

Detect key events in a football match like goal, penalty, corner kick, throw in, red card, yellow card, and substitution.

 
soccer model 2.png

Custom Face Model

Train a custom model on faces and recognize key players in the match or interviews.

 
soccer model 3.png

Visual Tags

Identify important scenes in the sports match or interview using visual tags & descriptions.

 
soccer model 4.png

Identify sponsorship

Train and detect logos and known graphics to identify sponsorship within the asset. See a summary of the total amount of time each brand is onscreen during the match.

 

Once detected, sport metadata can be viewed within the GrayMeta Curio extension for Adobe Premiere. This allows Editors to click and drag relevant scenes and clips onto a timeline, greatly speeding up the highlights or production process.


VALUE

  • Automatically create timecode specific metadata for the archive based on a range of machine learning service types (speech to text, OCR, facial recognition) from best of breed service providers, including GrayMeta.

  • Share metadata with existing systems, such as Media Asset Management, Workflow Orchestration and Editing, using GrayMeta’s APIs.

  • Speed up highlights and production processes using the GrayMeta Curio extension for Adobe Premiere

 
 
hussain-ibrahim-gZoRGh-GGKY-unsplash.jpg