Telepresence is to teleconferencing as dining is to eating. One is a mechanical process, the other transforms the simple act into a complete social experience. Or so the hope goes.
Also known as teeping, the idea is to create a completely immersive environment where you forget that the person you’re talking with is not physically present. Cisco is pushing this technology hard, but hasn’t crossed the line from teleconference to telepresence.
Teeping opportunity
I spent some time in a Cisco telepresence demo room this week, during a small business symposium co-hosted by the TIA and Cisco. It’s a cool system that will eventually lead to a true teeping solution.
On the plus side, Cisco has optimized the mechanics. Camera and screen placement, conference table set up, lighting and audio are all dialed in. You can sit down and look across your table at people sitting on what looks like the other side of the room.
Part of the trick is the way the physical layout adds the illusion of depth to the flat images on the large high def screens. In a few years, 3D video technology will make it seem spooky real, but the current system gives your brain sufficient cues to start filling in the missing dimension.
The final, great hurdle is enabling two-way eye contact. Until that’s possible, it’ll be teleconferencing, not telepresence. Right now, you have a choice: look into the camera, or look into the other person’s eyes.
The current iteration lets you see body language, which is a huge step forward. But our brains are hardwired for eye contact, and you can’t connect person to person with a stranger without it.
Case in point: when I’m riding my bicycle in traffic and I want to make sure a driver sees me, I look right into his eyes. We can both be wearing sunglasses — it doesn’t matter. Our primitive, hunter-gatherer brains instantly grasp the presence of a fellow human and go into “friend or foe” mode. It’s the same whether you’re running across the savannah or sitting in a corporate meeting. A split second of two-way eye contact determines whether you’re going to share lunch or be lunch.
A solution to this problem starts with some kind of eyeball tracking system, which determines where each participants’ eyes are focused on the screen. Software would then manipulate each individual’s image so that people on the other side of the conversation accurately perceive that individual’s gaze.
This solution requires huge computational capacity and magic software, rather than raw bandwidth, so Cisco won’t solve it. But Cisco and any other aspiring teeping vendor will snap it up in an eye-blink. So who has the chops to do it?
At this month’s Santa Cruz New Tech MeetUp, two of the presenters discussed exactly this kind of image manipulation. Pixim does real-time enhancement of video feeds, mostly for security applications at this point, and Pelican Imaging is developing computational cameras that can manipulate static, 2D images through three dimensions. The event’s sponsor, Santa Cruz Imaging, is also actively developing technology in this space.
In five, maybe ten years, brute force corporate R&D will solve this problem. Until then, it’s a genuine geek opportunity.