My phone recently received the upgrade to Froyo (Android 2.2) and as with many of the included apps, the Voice Search received a much needed upgrade. For example, I can now say "Send text to... John Smith... Hey can we change our meeting time to eleven [question mark] I'm running late [period]." The phone then asks me to confirm and then the text message is sent. The cool thing is that the voice recognition is good enough to understand most queries even if spoken in a casual tone (though I can do a pretty good robot impression to get it to understand the more obscure words).
This has me thinking of the possibilities of voice recognition software and, specifically, of the characters on Star Trek who could spontaneously say "Computer, what is our current trajectory?" or "Computer, how many dialects are there of Klingon and Romulan languages, respectively?" The computer would then respond quickly and accurately with an answer, in a crowded and noisy environment. This always seemed like, well, science fiction to me. But with recent phones like the Droid X and iPhone 4, I had an idea of how such technology could be possible.
The newest phones have 2-3 extra microphones just for noise-canceling purposes. Imagine if you had a dozen microphones in each room of a house ambiently listening, then you spoke a nonsense name which activates the query process. The computer then subtracts out the background noise, thanks to the other microphones, and performs the task or answers the question. If all the electronics in the house are connected through the voice recognition computer, then one could say such things as:
"[random name], turn on TV to channel 34 (or even just 'ESPN')." or
"[random name], play music by Frank Sinatra." or
"[random name], lower blinds in the living room." or
"[random name], lock the front door." or
"[random name], what is today's local weather?"
The future could be pretty cool.
No comments:
Post a Comment