Category Archives: Augmented Reality

The Coming Augmented Reality Language Semantic Conundrum….


 


A practical obstacle to workable mobile augmented reality technology is currently largely being overlooked.

From a recent Google patent abstract… :

“The method (augmented reality in the larger sense) may also include transmitting a query of the user to the server computer system to initiate a search of the history or real-world experiences, and receiving results relevant to the query that include data indicative of the media data in the history of real-world experience”

In the Google Project Glass video above, the augmented reality wearers make several verbal statements to be picked up by the AR technology for downstream processing. One character in the video uses Glass to translate a phrase into Thai. Another uses Glass to look up facts about a jellyfish, and another uses Glass to get directions while biking. Some exact excerpts are:

 “record a video…”
“hangout with the flying club…”
“take a picture…”

Each of these verbal statements can be made in many different ways by different AR users, without losing semantic (statement meaning) accuracy :

    “Get a video here..”
    “Ref the flying club..”
    “Hangout with my fly buddies…”
    “Grab a pic…”
    “Snap this…”
    “Take a photo..”   

For a simple expression like “hangout at the flying club…” we have 5 non-specific words that can be replaced with semantic equivalent word groups without statement meaning change. By just replacing each of these words with – let’s say – ten alternatives we have over a hundred thousand statements semantically equivalent to our original statement – and instructing the downstream AR in the same way as the original. For all the above statements in our example video, a very reasonable semantic expansion can easily run into trillions of semantically equivalent language inputs.

For any AR system to language-wise “equalise” all input semantic alternatives to a standard “base” that can for instance hit keywords in a programmatic backend environment, is impossible without near real-time massive input expansion.

Gatfol supplies this in milliseconds. With Gatfol, any AR input statement is semantically “amplified” in-stream (or simplified to the generic base) to make downstream language processing permutation-wise possible.

Use Gatfol to enhance communication in augmented reality…