Technology

AN INDUSTRY IN TRANSITION

In a rapidly evolving industry, we find out whether a holy grail of technologies exists that companies need to keep pace with, in order to lead the M&E bandwagon

By Anisha Gakhar

The media and entertainment industry is living by a zealous attitude of invent-or-go-home, and rightly so. This is evident with the constant innovations that can be witnessed in OTT as well as media intelligence. However, how are these technological innovations affecting business models – that was one of the topics that was discussed at the Amazon Web Services (AWS) Summit held at Mumbai in May, 2018.

ADAPTING TO CHANGING CONSUMER PATTERNS
Business models are maturing rapidly, which makes it imperative for media organisations to deliver more, seamlessly. Three attributes that help propel the goodwill of a media company are, speed, automation and efficiency. According to ABI research, OTT makes up 15% of the industry revenues, and is growing at almost ten times the pace of pay TV. Also, five years down the line, it could comprise one third of the market. In a couple of years, half the content on the web is expected to be devoted to streaming media. According to a report by The Guardian, British Broadcasting Corporation (BBC) hustled with 1000 extra hours of live sports coverage in November, 2017. Given this fact, it is important for media businesses to enter the world of streaming media as it is more convenient than reading walls of text.
At the Summit, Anil Nair, AWS, iterated the innovations for content distribution using the Edge. The diversity in viewership can be demonstrated by the fact that some viewers watch content at their convenience from the comfort of their homes or whilst commuting, and some, without electricity in remote locations. A part of the audience has Internet connectivity at their disposal and binge watch sitcoms over weekends. Some, like those in hostels, have limited access over the local WiFi. The array of viewership patterns is large and media houses need to be adept with these diversities, in order to satiate all viewership patterns.

An increasing subscriber base can cause bottlenecks at the origin server.

CLOUD-BASED OTT – THE GAME-CHANGER
Consumption of videos has increased in the government sectors, Video-on-Demand (VOD) sports, music events and gaming. The time taken to test, deliver and market new approaches, should ideally be minimal, to ensure immediate access to viewers. This is achievable through Over-The-Top (OTT), which allows the media company to sell audio, video, and other media services directly to the consumer over the internet by streaming content as a standalone product, bypassing telecommunications, cable or broadcast television service providers that traditionally act as controllers or distributors. The operators can monetise their OTT by hosting content on the cloud, and strategising their costs, time and avoid on-boarding premium OTT processes. Since most of the functionalities can be processed over the cloud, the device’s CPU resources are freed up, improving flexibility, performance and efficiency.
At the AWS Summit, Akash Saxena, head, technology, Hotstar, stressed on the need to use OTT, owing to the diverse demographics in India. Since, the one-size-fits-all strategy doesn’t apply to the subjective needs of multiple viewers; the need of the hour is versatility.
Another aspect to keep in mind is that live entertainment mainly revolves around sports. According to the discussions at AWS Summit, the NFL Thursday Night Football (TNF) was seen by an audience of 16 million viewers across all platforms. Amazon Prime Video reached a total of 1.5 million combined viewers worldwide in Week 4, and 1.7 million in Week 5. It reaches customers in 187 countries, some of which include, USA, Mexico recording the highest number of streams, Germany, a few Asian countries, South Africa and Tanzania.
Amazon Prime members watched the live stream on average for 55 minutes, which is a reasonable time span. Also, it is note-worthy that that the most popular alternate audio option was UK English. In order to ace the live-streaming quest, media houses need to ensure that their broadcast quality is top-notch, along with a world-wide distribution network. The system needs to be redundant and scalable. Personalised advertisements also generate exorbitant revenues, thus enabling a foolproof plan.

IT’S RAINING CLOUD
Mark Cousins, AWS Elemental, explained the media infrastructure at the AWS Mumbai Summit, 2018; emphasising on the usage of cloud technology. The cloud has gained popularity owing to its many benefits like paying and processing for content on-the-go. The trade capital expense is directly proportional to the variable expense, thus, making cloud a smart choice. It also increases the speed and agility, shortening the time-to-market in turn. This gives a marginal frame of time to test out new approaches. It also helps evaluate needs and adapt storage accordingly, resulting in benefits from the sale economy.
With cloud-based OTT, content can be streamed in multiple formats across multiple band-widths, and on various devices like TVs, laptops and smart-phones. One can simplify their OTT service deployment by installing an in-house Digital Rights Management (DRM) solution, or opt for OTT services that come with a built-in Multi-DRM. An increasing subscriber base can cause bottle-necks at the origin server. A pre-fragmentation approach, where all media fragments are harvested from the origin server and stored in the cloud, solves this problem. This helps bypass the dynamic creation of files when content is required. Instead, all files are prepared in advance and are ready to be delivered to the viewers, almost instantly.
The cloud can be deployed to be available globally, and instantly. In a matter of a couple of minutes, the content can go on-air live, across the world, with minimal lag. Using the cloud comes with a host of other pros, which can be leveraged to ensure a no-glitch set-up. The cloud is especially efficient at handling unpredictable bursty requirements, and choppy networks. This helps focus resources on the media needs instead of having to maintain data-centres.

MEDIA INTELLIGENCE – THE NEXT BIG THING?

For data driven image colourisation, systems uses AI to colourise a gray-scale image, thus, providing the capability for generating multiple plausible colourised versions.

According to a Netflix blog, over 75% of what people watch comes from recommendations. Personalisation is taking over the industry rampantly and is the basis of all push notifications, advertisements and custom data. The media industry has evolved from a singular status of content being available anywhere, on any device, at any given time; to, all content belonging to diverse genres being available instantly, across all devices directly to the consumer on-demand. |In earlier times, the storage and management of content was the main focus of companies, unlike today’s concern of what appeals to the consumer and their engagement.
Media houses are opting for fully integrated Artificial Intelligence (AI) and Machine Learning (ML) technologies to unlock the value of their content and data lakes. AI is the engineering of making intelligent machines and programs. It predominantly came to notice in the 1970s and 80s. It was followed by ML in the 1990s and is still used in most parts of the industry.
The ability of the system to learn without being explicitly programmed, is what makes this technology widely accommodating. 2010 witnessed the concept of Deep Learning, which is a sub-field of ML, and is concerned with algorithms inspired by the structure and function of the brain, called artificial neural networks.

ML AIDED CONTENT CREATION
Can content ever create itself? ML aided content creation involves many factors, like colour correction, scene extraction, highlights creation, automated editing, conformance/compliance and performance optimisation. Caption generation for photos and videos, translations in a set of languages and automatic narration generation are also possible with ML. Graphical functionalities like roto/depth detection, 3D conversions aid in enhancing the quality of the output. Additionally, real-time lip sync of CG models and VFX bidding is also supported by ML. Sub-title creation and its translation is auto-aided by ML.
For data driven image colourisation, the system uses AI to colourise a grey-scale image, thus, providing the capability for generating multiple plausible colourised versions. Face recognition through automated meta-data tagging and extraction, person-in-scene detection and person-in-picture tracking are all feasible with the use of intelligence. It also facilitates logo detection, text extraction and speech-to-text conversion. The person-in-scene tracking is especially useful during sports highlights, for instance, tracking the person who hit the goal during a football match.
Another aspect of ML is contextual search and advertisements, which uses the user’s age, emotion, story, geographical region, content execution and character to generate content, and eventually TRP.
The inputs to ML are sentiment analysis along with viewing patterns of consumers. These are applied for content discoverability, recommendations, personalised dynamic ad-insertions and content production investment decisions. In summary, there is no best-way to execute ML in a business. Yet, consider best-of-the-breed toolset for each task. It is important to not get swamped by just platform in a ‘box’.
Integrating the content lake with ML, and the security (of content lake as well as means by which service providers/applications access content) holds utmost importance. Also, selection of the cloud platform can be chosen on the basis of the eco-system on which the company thrives. The workflow and the flywheel best suited for one’s business, should be built in accordance with the company’s requirements and priorities, with augmented ‘human’ element being of key essence.

HELLO, TOMORROW
The future belongs to those who are swift at accommodating new fads, and getting done with redundant technologies. As organisations in every sector clamber to deploy ML in practical instances, the big players have been taking advantage of advanced algorithms and personalisation for some time now. AI is the answer to some grim challenges hurled at the media industry, since it can be used to analyse the appetite of the target audience. While it looks like the next big revolution, accompanying this technology with cloud services, is a double-whammy for a media house, and a sure road to success. Realise. Re-invent.

Previous ArticleNext Article

Leave a Reply

Your email address will not be published. Required fields are marked *