Calm down, the robots won’t take your video editing job. You’ll just get a little helper
Here’s how cognitive computing can help editors make sports highlights
That’s the deal: in the sporting industry, the content is what gets you new viewers and keeps them coming back.
The stadium can only seat that many people, right?
So we get global with it: broadcast sporting events over radio, television, and the internet; post game scores in the newspapers and the websites; create media companies dedicated to covering all things sports.
That content includes the highlights as well — a sure way to get on that hype train during or after the big event. Post the dunk, strike, or touch down on TikTok — get a ton of views — see the viewers pour in for more.
The problem is — you’ve got to keep the production rolling at all times.
Human editors don’t cut [enough of] it
Most times, sports leagues and media companies covering sports events rely on human editors to cut their highlights.
That results in folks spending enormous amounts of time in the studio, going through the game footage and cutting the best highlight-worthy moments, then editing those into the compilation.
Now, say you have several videos to make. Those may be the compilation of three-point shots, quirky mascot dances, or individual highlights for certain players.
And that is just one game. How many events a given company covers mid-season?
At the end of the day, you get a lot of time dedicated towards somewhat routine and mundane work, which could’ve been spent more productively on creative tasks.
Artificial Intelligence: here’s your little helper
Addressing the issue of wasting time, some look into automating the process with Artificial Intelligence.
Instead of sitting through each and every match, cutting the clips, and sharing those over many platforms for your audience to enjoy, you can now have AI do that for you… In a limited set of use cases, though.
The thing is that sports highlights are all about the context. How does an algorithm tell an own goal apart from a regular one? Can it differentiate between low and high-impact plays?
The short answer is: not really. Most of the AI models are built around simple object tracking and deep learning algorithms, and require additional data from external data sources, like audio analysis or tags. The tradeoff here is not very beneficial: yes, you don’t have to cut the highlights yourself, but you do need an external data source, and spend some time training the model.
Apart from time, you will also be spending more money: getting an external data source to support the AI model is often very expensive. So much for optimizing the effectiveness of the whole operation.
Cognitive computing: here’s your better little helper
To do that video analysis thing effectively, we have to go deeper than simple object recognition. That is where cognitive computing technology can help us.
As we have established, picking up the context is essential to identifying highlight-worthy moments. That would require the thorough analysis of visual clues and informed decision-making based on incomplete sets of information.
Sounds like something only the human brain would be capable of, if anything.
Or… how about a robot with cognitive abilities close to those of a human?
Well, at least in terms of video content analysis.
A solution powered by cognitive computing technology is miles ahead of traditional deep learning models for analyzing large volumes of footage and identifying things in there. Because of the tech under the hood, it needs no additional data other than the video itself: no meta tags, no timecodes scouting.
We do analyze the audio sometimes, but only to make sure that the commentary does not get cut off and sounds consistent.
Using the high-tech cocktail of deep learning models, cognitive and traditional computer vision plus digital image processing, the software can actually “see” like an actual human.
That allows it to accurately analyze the footage and pin-point highlights of the game.
That poses the question: will media companies kick their editors out the door and use the technology to completely automate the content production process?
So, let’s talk about it.
Cognitive Computing augmenting human effort at sports highlights generation
For all impatient folks out there, here’s your answer: no, the robots won’t replace you at the job. It will just spare you from tedious and time-consuming tasks.
And now, onto the longer version of the answer.
Artificial Intelligence sure won’t do it. Its lack of agility and precision requires a human constantly moderating the results it puts out. So, there’s not much automation going on.
You were wasting time cutting the highlights yourself, and now you are wasting time watching the AI cutting trailers.
But what about cognitive computing?
Given that it has human-like cognitive capabilities, it must be able to replace us in content creation, right?
Not really.
Sure, the solution is great at analyzing video content: the use of cognitive science, computer perception, and probabilistic AI allow it to effectively extract important data from the footage.
And that’s not just tags and scores — you can use the software to automatically generate your sports highlights, and push them to social media platforms.
That is very beneficial for marketing and content strategies that media companies, sports leagues, and broadcasters carry out, but we are forgetting the thing that is behind all of that.
And that’s human creativity.
Creativity is the concept so complex and deeply rooted in our psyche that no tech of today and near future can replicate that.
Instead of following that trail, technology offers something a lot more manageable: a way to minimize the human effort dedicated to mundane work, so that content creators can spend more of it on — you know — creating the content.
Humans get to come up with effective marketing strategies, generate ideas for new content, and be out covering everything there is to a sporting industry.
They get to broadcast the events, make sports-related shows, interview players, cover press conferences, write analytical articles, and anything else they can think of to make their brand more recognizable.
And at the back-end of it all, there is a cognitive computing tool that handles more tedious stuff. Feed it the game footage after the event is done and get back to your work. It will analyze the video, find the most important moments, and edit them into a compilation.
Then the editors can take over to add some of that creative jazz to the mix. Tinker with transitions, create and add motion graphics, experiment with music and sound effects — everything you need to produce the best sports content out there. Just without the tedious work of picking the footage apart.
And it’s not all.
It is a no slouch when it comes to customization. Need a players’ brawl compilation? Not a problem! Want to get a clip of the best moments’ clip for a particular player? You got it, boss.
Then there are things that make the software one step ahead of you.
Cognitive Computing can do much more than editing clips — thanks to its video analysis capabilities, it can perform other data-driven tasks like:
- Cropping the footage from landscape to portrait aspect ratio;
- Filtering out the inappropriate content;
- Inserting relevant ads based on footage context;
- And more.
Put your effort where it matters most
Technology like cognitive computing provides a way for media companies to augment their content-related efforts.
The software does a great job at handling video content analysis and sports highlights generation, leaving the editors more time to focus on more creative projects. So the editor gets a better task, fans get their highlights to trade, business owners get their metrics.
Everyone is happy.
Schedule a DDIChat Session in Digital and Content Marketing:
Apply to be a DDIChat Expert here.
Work with DDI: https://datadriveninvestor.com/collaborate
Subscribe to DDIntel here.
