Post Top Ad

Your Ad Spot

Friday, 13 April 2018

Voice Recognition | Siri | Alexa | Cortana| Google Assistant | Bixby !!

Most of you would have talked to their phone or seen commercials showing how smart the technology is where you don't even need to touch your phone to get the things done, all you have to do is just say- OK Google, Hey Siri, Hey Cortana. 

Image Courtesy: Google

Well, today's blog is all about how this speech recognition system works. And then we will have a short comparison between the most used ones- Bixby, Cortana,Alexa and Siri.
How simple it seems to look - our phone is smart we just talk to them and they reply. But you can't even imagine how complicated it is. It seems like its easy to just add words from the dictionary to the computer and if people say those words then the relevant action is performed. It's not this simple how it looks. If we talk about it's evolution and how it has managed to be this much better is really appreciable. At first the voice recognition system was just for few specific commands and the user had to be very careful while saying those words in order to activate that command also the assent of people matters. Some speak the same English word differently while others differently so in order to understand what has been said is really a tough task. There are so many words that are similar in fact some are same in their pronunciation like sea or see or c. Now to judge which C has been said is really a tough task. There are so many algorithms working just for those small things that for us seems to look small but are really complicated when it comes to implementation. This is why we need an internet connectivity for most of our works as the match is to be done with the databases stored on servers.
There is cloud that comes in here, because the questions that we ask aren't predictable and is done with the help of vast databases stored on those servers and after processing they use to understand what we want.

Previously when it was introduced we had to be a lot careful while saying but now it comes natural. You don't need to change your tone or assent. Just speak the way you speak and the speech recognition system is there to understand whatever you say. It is now that they have been a lot more better in all terms whether it is for providing input or getting they output rapid fast. The comparison is also cool as they are getting smarter day by day.  
It was first started in iPhone's as Siri.  
Let's understand it with a example,
Suppose that you searched for-
Who is the 45th President of United States ?
Surely you will get the answer as Donald Trump is the 45th president of United States. But if the next ques asked is- what is his age? Here his refers to Donald Trump but we are not actually saying it.
Here we are not specifically asking for Trump's age we are asking the same ques differently. This is what it must be capable to understand. And if it is capable of this then its really smart.
In terms of searching I don't think anyone could beat Google whether its Bixby from SAMSUNG, Siri from Apple or Alexa from Amazon. A few days back there were buzz of Apple hiring employees to make their Siri better. 
If we compare the devices in terms of  direct questions almost all the speech recognition work well. But if the query is a bit complex like asking Donald Trump's age indirectly, is where it need to get smarter. And Google Assistant is far better in this. Well the battle between others and Google Assistant is head to head but still according to me Google Assistant has an upper hand.
The best feature that I like about Google Assistant is- it remembers everything you have asked it to. I am use to forgetting things. So all I need to do is just ask Google to remember those things for you and then you can simply ask Google if you said anything to remember. All the stuff just by saying OK Google..Really Cool..!
Thank You..!!

No comments:

Post a Comment