Here's how media and entertainment industry is leveraging artificial intelligence to transform the space

Artificial Intelligence (AI) is not only gaining prominence in sectors like healthcare and education, but also impacting the media and entertainment industry. From content processing and recognition to speech recognition and machine learning (ML), there are myriad ways in which AI and ML are gaining traction in the sector.

Talking to Moneycontrol, Girish Menon, Partner and Head, Media and Entertainment at KPMG India, said, “AI is definitely being used in India. In the past year or so, we started to see a fair amount of traction of AI usage in the backend.”

“One of the areas, for example, where AI is being used significantly is meta data generation. For example, if you have a video clip, in a digital environment the more data you have about that video clip, the greater the ability to use that video clip in different situations. Many elements that are descriptive of a particular scene which you are able to tag, based on which you start creating recommendation engines in a far more efficient manner than you can do on a manual basis. Where AI comes in is there are AI engines and software and there are few startups that are trying to specialize in this field which then automatically identify specific elements within a particular scene once you run that scene through an AI engine. This is one specific area which is seeing a lot of value when it comes to AI,” explained Menon.

The other aspect where AI is used a lot in India is embedded with data analytics, said Menon.

Data is being used as a powerful tool for marketing to audiences. Sentiment analysis is being used to target territories with high preferences, and screen count across territories is being determined based on social media analysis. Film scheduling on television is also being done based on viewership data and digital uptake is determining which films to invest in for OTT (over the top) platforms.

“When you talk about digital user base and digital statistics that’s when AI comes into play because you can analyse the data through AI engines and start making predictive analytics like what a user likes to watch or what kind of content will work all of that is getting a more science behind it through AI,” added Menon.  

Online video platforms are using AI aggressively to personalise content for every individual. Platforms using AI have algorithms that learn over time what users want to watch and recommend similar content, including when the recommendations should be made for most effectiveness. Increased personalisation results in increased time spent as well as engagement and hence it is becoming a must-have feature for all platforms of scale.

A point in case is Netflix which has adopted an open source approach towards artificial intelligence and machine learning. The platform uses algorithms to adapt the entire user experience to each individual subscriber.

Whether Indian OTT players are making use of AI, Menon says, “Indian OTT players are using data but is the data science or the AI engine or the recommendation as sophisticated as a Netflix probably not but Netflix has taken many years to get where they are. You have to give these guys some time to reach there but everybody is aware of the benefits of AI and everybody is investing in those areas.”

Apart from OTT platforms even Bollywood has resorted to AI to maximise reach of films.

An EY 2018 report on media and entertainment elucidates this point with an example. Producers of 2017 release Shaadi Mai Zaroor Aana collaborated with Infinite Analytics Nucleus to apply analytics to pull audiences for the film. Infinite Analytics used its artificial intelligence (AI) platform called ian (Infinite Artificial Nucleus) which offers product recommendation system and represents it across all channels targeting specific users to maximise reach of films. This was the first time analytics was being used to pull audiences for a Bollywood film.

Infinite Analytics identified the lead actor Rajkumar Rao’s films, music, food, books and recommended the same for the film. As a result of successful implementation of data analytics, the trailer of the film reached a million hits within five days of its release. It is likely that this trend shall only grow in the years to come.

And that’s not it. Companies are training machine learning algorithms to help develop film and television promos. The process begins with researchers feeding the system with multiple film trailers. This is followed by the AI system being trained to conduct content analysis tagging scenes from similar movies according to different emotions, tone of voice, musical score based on the feelings invoked and other aspects of a film such as location, lighting and framing. After the ML process, the movie has to be put into the system so that the AI could select which scenes would appeal the most to viewers.

In this process, the entire process of making a trailer takes about 24 to 48 hours compared to the traditional process that could last up to two weeks.

When it comes to novel ways of marketing films, Bollywood has never shied away from trying new things. Baaghi 2, one of the successful films of 2018, introduced a chatbot at the end of the film’s promo allowing people to strike a conversation with the film online.

Experts say that although use of AI is an emerging phenomenon, it will soon become mainstream.

Pointing out the other area where more and more AI action is happening Menon said, “Localisation which is something that is happening globally. For example, you need to convert a particular content into a different language, if you need subtitles, if you need dubbing, all of that is now for example AI engines are being embedded within that to make the conversions more automated. The initial versions can be done through AI and then you can fine tune it manually.

“What I have heard is that that Google is trying to get into real time translation automatically through AI where if you want to watch a YouTube video it may be shot in English but you can get an automated Hindi subtitle as it plays,” he further said.

“One other area where you will AI happening more and more is actually voice search because that’s one area where most people are investing a lot of money because as we go into more regional language and more non-English speaking audiences so access is a critical element so that’s again where AI is used at the backend to make the machine far more global when it comes to recognizing languages Amazon has Alexa, we got siri,” Menon added.

AI is also being developed to enable creation of immediate and quick highlights of sporting events and capture metrics to enable statistics and scoreboards automatically. This is done by combining a review of on-screen action, historical data, crowd reactions, Twitter posts and players’ facial expressions wherein AI systems identify and rank the best shots that made up the day’s highlights.

However, Menon said the elements of AI that are being used in the media industry in India are not mainstream as yet.