A Video in a Haystack: Online Video Industry’s Next Big Challenge
From the early 20's, when the film industry introduced synchronized audio and motion pictures to the world, Video technology has undergone many changes, faced different challenges and we, the video consumers, experienced them all.
The current challenge has not even fully revealed its face to us, but we can start sensing its impact in our daily video usage.
Quality issues across the board were definitely a main challenge for decades and even not so long ago. It started as low frame-rate, black-and-white silent motion pictures that didn’t even capture real life as we experience them, and continued with slow and buffered video streaming over the internet in later years.
Nevertheless, display devices of today, like mobile phones and televisions, support extremely high resolution (4k of higher) and streaming technology and services over the internet gradually commoditize it and adopt that as a standard.
We've transitioned from ‘black and white 16 frames per second silent film at the cinema’, to a ‘HD video on my mobile phone while I'm at the park’.
Different video Formats have been a rather recent challenge for the digital video industry and for all of us as consumers. From the early 2000’s different vendors bombarded us with different video containers, encapsulating different video codecs, suitable for different devices, browsers and operating systems. Trying to playback a video on a screen didn’t always end up with success.
Nowadays, vendors are consolidating and standards are pretty well defined across our multiple number of screens.
Recording and Storage of video is definitely another challenge the industry and its users were dealing with. Video recorders were sold for $50,000 in 1956, and videotapes cost $300 per one-hour reel (1).
In 1979, 1 megabyte of digital storage would cost 233$, in 1995 it cost 0.75$ and in 2005 the same megabyte cost just 0.0006$ (2).
While in the past recording video was expensive and you had to think twice before doing it, today, all the video footage you own, probably doesn’t cost more than your weekly grocery.
That brings us to current times, forward looking to the future and the changes yet to come.
The video footage we will own will increase rapidly - both organizational video and user-generated content. It is estimated that in 2017 there will be 4 billion online video users (doubled since 2012) (3), 80% of all consumer Internet traffic will be video by 2018 (3) and 58% of all business Internet traffic will be video by 2017 (up from 31% in 2012) (4).
When this happens, Searching for a specific part in an endless pool of content will be a daunting mission, an impossible one even. That’s the industry's next big challenge.
Searchability is a key factor, nothing surprising about that. We’ve been seeing technologies starting to deal with that for some time already - Speech-to-text, Optical Character Recognition (OCR), Voice or Face or Object Recognition while technologies like Big Data or advanced Semantic Search Engines support them.
What will change in the future is the magnitude – the volume of video around us will raise the bar of how complicated it will be to search in it.
Look at some of the disruptive technologies that start to emerge today and how they naturally encapsulate video in their use-case.
Drones which are already flying the sky, will become popular for many use-cases (security, commerce, leisure are just few examples) and will capture video from every possible angle while flying to fulfill their mission.
Wearables (Google glass as a harbinger) will record video of our ‘point-of-view’ while providing information on what we see.
Everyone eventually will have Smart Homes with a network of video cameras and the cities we live in will be camera-netted as well.
Our Transportation will record the road and the cabins 100% of the time.
Advanced video conferencing, with or without holograms, will reduce our need to travel and meet our colleagues and love-ones.
These and more have started already to appear and there is no reason to think that the trend will all of a sudden stop or shift.
So it is not that video itself is the future’s technology but it is simply a native part of so many other technologies, services and their use-cases.
With your permission, let’s take it a couple of steps forward, Sci-Fi style, to the times that these technologies will really shine and mature. Without effective, quick and elegant ‘video search’ we simply won't be able to get along. How come?!
Imagine a video revolution so drastic that our children and grandchildren won't even need to remember things like we do, as more and more events will simply be there to re-view...
More and more of their lives and actions will be recorded on video – from the smart-homes they’ll live in, the cars they’ll drive, the smart-clothes they’ll wear or the camera contact-lenses in their eyes.
Their world would be a different place than ours, and not because of flying cars or humanoid robots. Imagine crime and law enforcement relationship in this filmed-type future. How an eye-witness’s evidence in a trial would be like. Not only would their words be heard in court but their actual “point-of-view” taken from their personal cameras could be presented.
Imagine the way this generation will be doing Business, Politics or going to university… Having their entire meetings, classes, formal and non-formal interactions optionally recorded on disk and basically every point of view can be reviewed.
To us it sounds like Internet and Google would sound for a 17th century scholar.
So, in such a futuristic world, or even one similar to ours, how do you search for a specific word, sentence or event, in years and decades of recorded data?
The video and big-data industries has been working with some of the mentioned technologies, but it shouldn’t stop there, not with the future we just painted. Much more ground needs to be covered to deal with the volume and information-retrieval users will require. Video footage itself will have to be "smarter" ,the tools that capture it, the systems that process it, tangent technologies to it (Internet-of-Things for example) and obviously us - the viewers.
Exciting times and enormous challenge ahead of us.
(1) Elen, Richard. "TV Technology 10. Roll VTR"
(3) 2014 Cisco Visual Networking Visual Index Forecast