Voice Recognition:Explained!

Voice Recognition, a technology like VR where someone is always claiming that they have made a big break through and it still turns out to be some kind of crap. The history of Voice Recognition is pathetic and we do not mind it beside its success today and finally it is gaining widespread adoption. The early forms of voice recognition had very limited vocabularies, some systems only recognize about ten words and even couple of decades later that number had grown to only around 20,000 which may seem a lot but remembering that the English language has over 1,000,000 words it feels miserable, isn’t it? On top of that early software could not predict what words you were trying to say by using context so these programs were even worst. Fortunately we finally got computers in the late nineteen eighties and early nineties that had more storage and processing power allowing them to comprehend more natural speech instead of forcing you to talk like the way in which you seemed weird.

Nowadays technologies like Siri and Cortana don’t rely on a limited dictionary or the relatively weak processor in your device at all instead they use huge cloud database that store millions of words and phrases and have lightning fast CPU to understand what you mean with much more precision. Google Speech recognition software even learns from real search engine and can also recognize a variety of accents so you can use it, whether you’re from Eastern Canada or Southern Texas but can the power of the cloud do more, when you ask your phone for a sports score how does it know to tell you how your favorite team is doing instead of giving the list of sports store. Well, most of the present voice recognition does not listen one word at a time instead it also listen to other words for context and uses probabilities to determine what you’re trying to say and this is a complex process that uses many difficult mathematical models. Google for example uses an artificial neural network that functions similarly to your own brain using digital neurons to learn what people are saying and you can actually see this I action as Google now often changes what it thinks on the fly as you continue speaking and it gets better.

Greater amounts of processing power have enabled everything from real time translation tool being able to talk to gain characters with a VR headset to in motion tracking in which a computer can use the timing and pitch of your voice to figure out how you’re feeling. Furthermore it is even deployed in the fighter aircraft so pilots can concentrate on mission objectives instead of filling with cockpit switches but also those recognition has come a really long way its proliferation has presented us with some new challenges, one big concern has been finding ways to filter out background noise so you’ll still get correct results even if you’re standing in the middle of a busy street or speaking of standing in public. Another massive issue is privacy. Many types of voice recognition software improve upon themselves by learning user habits and combined with cloud processing, you’ve already seen some real concerns such as with Samsung smart TVs earlier this year which had a privacy policy which some people believed allowed Samsung to monitor your living room conversations which indeed is unacceptable.

Keeping all the cons a side voice recognition is really cool technology that can indeed the impress anyone let us hope this tech last with continuous development so that we can have our own artificial assistant in near future.