From Augmented Reality & Computer Vision to Seamless Interfaces


We all hear a lot of noise coming from the buzz words Augmented Reality (AR) and Computer Vision. it seams like every company making more than $10M a year wants to create an AR app to do something apparently game changing for their business. Every tech company in the world has vacant position for Computer Vision Engineer. But why? Is this just a fade, some over hype phase or a true deep movement toward something big?

TLDR :

  1. Yes there is Hype on AR and Computer Vision. A lot of noise for useless apps, imaginary hardware or invisible prototypes.
  2. But there is a real deeper transformation and evolution on what we can do with recent technology.
  3. All that is leading toward the next generation of computing : seamless interfaces

The Hype

Yup, there is no other way to put it. There is a lot of hype about Augmented Reality. Think about Google backed Magic Leap : 2 Billions Dollars in. And no official demo’d prototype. It’s way more hype than A$AP Rocky making a Dior Campaign. Banks creating corners for Hololens in their field agencies. Yup, that also is pure hype. A lot of what we see and hear about Augmented Reality is noise. It lacks concrete application, business opportunities or real technology.

But what did you expect? Augmented Reality is all the sci-fi industry have been selling us for the past 3 decades. You have examples EVERYWHERE. Literally any movie wanted to showcase a futuristic forward looking computer interface. Some of them included what we call AR today. Robocop’s glasses, StarWarss Holograms, IronMan’s 3D awesome interface. We were ready to go nuts with Augmented Reality by being exposed to it without being able to access it. And now, we get to test and buy the first real-life version of it.

Same thing for Computer Vision. Since AR needs it even for the most basic use of it. Even for simulated AR like the one you can see in Pokemon-GO, (where the user is looking at real-life objects + images generated by the computer on his/her screen), complex computer Vision is involved. On top of that, every autonomous car business is hiring Computer Vision engineers like crazy. There are thousands of direct down to earth industrial applications for it. It’s the perfect storm to make Computer Vision one of the most hyped tech topic these days.

The Real Deal

But here is the deal : It’s more than just pure hype. Together Augmented Reality and Computer Vision represent the biggest opportunity for a massive shift in the tech industry. A lot of investments are happening in that field, from VCs as well as internally in larger corporations. Apple has filed for patents around glasses. Microsoft has made a huge PR campaign with Hololens. Google Lenses is coming, Tango is up and running, and Facebook has its own stake in the game as well.

Let’s look a bit at deeper layers of this trend. The reduction of the computational power required for standard AR operation is the key to reducing the cost of the hardware. Big names have been working on that, Intel acquiring Movidius was a important move in the Computer Vision field. Their chip is game changing for visual sensors. Still at an earlier stage, the startup 8th Wall works on the software side of things, to bring AR Apps to 90% of the mobile market. All these efforts make leaps forward possible. They lead to better, cheaper, cooler Augmented Reality Hardware and Software. Soon everywhere for truly useful applications. Not unlike Artificial Intelligence. Computing power rises, algorithm optimisation allows sorting and categorising more efficiently. Hence a computer can tell the difference between a dog and a muffin.

Finally, the tech industry is on a path to connect everything. From your coffee machine to your lawn mower. And smart things are smart only if they have valuable interactions with their direct environment. Here comes Image Recognition for EVERYTHING. And I mean EVERYTHING. Even the next iPhone is supposed to unlock when it sees your friendly face. Mixing tech and real life a bit more everyday.

The true end game

I hear you say : “Ok, cool, so what?”. And you’d be right to be unbaffled. Because none of that is truly game-changing. It might be cool, but it is not as superbe as all the Human-Machine Interfaces the Pop Culture has promised us.

The final goal of that progress, investments and trials are tending toward is the next generation of computers. It won’t be about computing power. It won’t be about graphics or Apps, or bezel less screens. The next big thing in computing is implementing it seamlessly in the real world. Technology will be used without a thought, like a sheet of paper. When you will need a wall to become a tablet, it will. When you’ll need your desk to show you inspirational pictures, it will. And you want to change the volume of the music, you will just slide your finger on the side of your chair. Why? Because your seamless computer will remember you said out loud once, that you like your volume settings to be accessible at all times. Your bathroom mirror will display the news. Everything, everywhere will become the computing surface you need. And tech will fade in the background when you won’t need it. Walls, tables, mirrors, chairs and sofa will still be what they are. A layer of technology will be added, like magic, and disappear, exactly like magic.


Photo by Edho Pratama on Unsplash

What about design?

Designing seamless interfaces will represent a major challenge for UX / UI and product designers. For Human-Machine Interfaces (HMI) to be seamless, a lot of the interactions must happen outside of what we today consider as the interface. Think about your smartphone for a sec : All what you do with it happens through the buttons, the screen, the voice. For the next generation computing to be really seamless, a lot will happen the same way you interact with other people. With words, with obvious gestures, but also with a more “under the radar communication”. All the subtext of social interactions, from your body language, to the ton of your voice, the sparks in your eyes, etc. Hints, if you will. And the designers task will be to think of new way to get those hints, and design the resulting action. Computers will have to pick up those subtle hints to provide the best User Experience possible.

I might be completely wrong with all that seamless interfaces thing though. Maybe we’ll be happy with bigger screens and more ergonomic mouses. Yet, think of the logical evolution of HMI. From keyboard + Mouse + screen (you click somewhere to have a reaction elsewhere) to tablets and smartphones (you click where you want the reaction to happen) to seamless interfaces (the interface is where you need it to be, and how you want it to be). A bit more logic than going from smartphones to always on VR sets, isn’t it?

Finally, if I were to give a timeline on seamless interfaces, I would say 5 to 8 years ahead. Today, early prototypes, mixing projection and image recognition, Augmented Reality Apps and Hardware are going in every directions. But the original basis for seamless interfaces is there indeed. Projection hardware is still a bit clumsy to be fully ready. Image recognition softwares are getting there, probably faster than we think. And Artificial Intelligence is progressing at a very fast pace, and we can assume that it will be able to provide contextual interfaces while its sensors pick up hints.