A picture may be worth a thousand words, but computers have a very hard time doing something even a 2 or 3 year old child can easily do: recognizing what's what in a photo. But we're getting there...
After developing algorithms that could loosely categorize images like "sunset", "flowers", and "beaches", Google is now working on something that can more precisely describe what's on a image.
This system can not only recognize differente elements on a picture, but can also translate that into natural language, resulting in things like "a person riding a motorcycle in a dirt road", "a group of young people playing a game of frisbee", or even "a heard of elephants walking across a dry glass field". Sure... for each one of these correct descriptions there are lots more that aren't that accurate (or fail miserably or comically, depending on your point of view.)
In any case, it's simply a matter of time till these systems get more and more accurate, and will allow us to do things like searching for images, and movies, based on a vocal description ("show me movies where a couple is drinking wine by the fireplace") as well as giving a new sight to blind people: just imagine pointing your Google Glasses anywhere and being able to hear what you're looking at.
Sounds like science-fiction, but it may become science fact sooner that we think.
No comments:
Post a Comment