Blog posts related to Augmented Reality

HMD-less 3D

IEEE Spectrum has a very interesting article on HMD/Glass-less 3D systems on its December 2010 issue. You can read it here. The article presents past and recent efforts on creating 3D displays without using spectacle-type displays, discussing various techniques. We find the possibility of having 3D displays without the whole – arguably obtrusive – arrangement that comes with HMDs, quite fascinating. Some of the recent HMDs are indeed small and quite comfortable but at the end of the day we feel it would be more intuitive and… well, normal, not to wear them.

Panagiotis Ritsos with HMD

A not so normal pair of glasses

Nonetheless, as we understand it, the current systems work on a number of assumptions – such as distance from the screen, inter-ocular distance, posing certain usability restrictions.  Among the companies mentioned Nintendo, for example,  offers some adjustment for such variables in their prototypes in an effort to rectify these issues. Presumably an active (or real-time re-) calibration of the system could offer some transparency and independence to the user but that is a research subject on its own. Moreover, the techniques presented in the feature are governed – once again – by quality-cost-power consumption trade-offs, with obvious consequences on mobile usage such as portable game consoles or smart-phones. Still, the concepts are there and people keep pushing technology and that is the only way to go.

One aspect of 3D HMD displays, that the article comments on, is the ocular and non-ocular effects that often come as a result of HMD usage. In fact it is one of the aspects we on our past research efforts have examined to some extent. These issues, most often headaches and less often nausea and vomiting, are a true limitation of HMDs. Top that with with focus, luminance (for outdoor use), resolution, ghosting etc, things get messy quite fast. It is surprising – well maybe not so much if you want  to be cynical – that people do not mention these limitations when talking about ‘AR-hype’ . Arguably, the focus today is on handheld mobile AR, but nonetheless these limitations are present for years and researchers have investigated them to a large extent, particularly in the context of military simulators.

We know you can not market something when you focus or mention its problems (the aforementioned cynicism…) . As UX researchers though, in a broad sense, we need to take these things into consideration if we want to produce truly immersive and commercially successful systems.

Photorealistic Rendering for AR

Probably the most impressive examples of AR I have seen in a while. Finnish VTT Team has done some impressive work in the past and this example is nothing sort of spectacular in my book. The video was recorded a Dell laptop with a Quadro FX 3700M video card, a Core Duo processor and a basic Logitech webcam. The graphics itself is drawn using OpenGL and GLSL.

VTT’s description of the above video is:

Photorealistic rendering for Augmented reality. Uses soft shadows, indirect lighting and image quality matching. Various materials like glass, chrome or plastic are possible. Lighting is automatically determined from a ping pong ball.

For us it is a demonstration how realistic things can appear with modern hardware – albeit not of the ‘handheld’ type – and GFX APIs, hinting that the more we push the envelop and exploit technological advancement, the more immersive things will appear.

Storytelling and AR

Kat Austen in New Scientist’s Culture lab presents a very interesting post on AR books. By AR books we – currently – refer to books that have some added information in the form of extra visuals which are projected through a a web-cam-equipped computer, tracking the book’s features, such as pictures, QR Codes etc. The post mentions  some of the books that appeared this year, such as Fairyland Magic from Carlton Books and Tyrone the Clean’o’saurus from Salariya Publishing and presents the author’s experience with them. Kat also mentions Camille Sherrer’s Souvenirs du Monde des Montagnes along with a video bellow.

We cannot help but agree with the author’s overall assessment that, overall, the current AR books have an element that makes them impressive initially but they eventually lose their grip on the reader – particularly the young ones. Moreover, the set-up required usually to achieve those augmentations is not exactly convenient, especially when you simply want to ‘dazzle’ your kid with an elaborate story.

It so happens that this is our view on many matters of current AR. Things are impressive on first sight but when you pass the ‘wow, what’s this!’  point, they often fall short. We are planning  to investigate in the future, the sense of immersion of various AR book paradigms through user assessments.

Total Immersion CEO on AR

Bruno Uzzan, CEO and Co-Founder of Total Immersion, speaks about the past, present and future of Total Immersion during his keynote at AR Immersion 2010, November 9, Los Angeles, CA. (Source – T.Immersion blog here)

Interesting points in our humble opinion are:

  1. The ‘cuckoo’ talk looks cool and dare I say kinda funny (Video1- 3:12)
  2. Mentioning crudely the impact CPUs (and dare we say GPUs) and 3D GFX APIs have on modern AR
  3. The need for camera vision for modern mobile AR
  4. That the quality of graphics affects immersion (ok, it is kind of implied)
  5. Wireless baud rates
  6. Smartphone worldwide usage and projections

We feel points 2,3 and 4 are of paramount importance of the current shape of handheld AR. In fact the quality of graphics and the effect this has on the sense of presence of the users has been a key part of our past research.

For more videos from AR Immersion, visit Total Immersion’s AR Immersion 2010 YouTube Channel.

Vuzix see-through goggles

Technology Review mentions today the new Vuzix Wrap920AR Goggles today, in the midst of current AR popularity. The Wrap920AR is equipped with two video cameras that ‘see’ the front of what the wearer sees and project it on two LCD video displays, resembling what is allegedly a 67-inch video display at ten feet. The have stereo capture, 6-DOF head tracking, VGA connectivity and come bundled with some nice software bits.

Wrap 920AR

Wrap 920AR

Now, their price is about 2K USD which is not much compared to similar camera-equipped see-through HMDs but it is way to much for me to get a chance to try them. So… I will attempt to imagine their functionality.

Joking aside, they seem light and less obtrusive than past implementations. I imagine it wont be long before researchers use them. The fact that what the user sees is essentially a video feed from the cameras could potentially provide some nice opportunities for advance synthesis virtual and real objects. Moreover, compared to optical displays they can tackle ambient light – as the article points out.