Another Level of Speech Reco...

October 7, 2011

After months of growing anticipation, Apple announced the release of the iPhone 4S on Tuesday No one was really sure what sort of iPhone Apple was going to release, but many people were banking on the iPhone 5. For weeks, tech blogs like Mashable and Wired have written extensively about the long-awaited release of what everyone assumed was going to be the iPhone 5. What would it’s functionality be, how would the screen, camera, and software be upgraded and improved?

To the dismay of many, Apple has given no definitive insight into when they will be releasing the iPhone 5, instead choosing to heavily promote the 4S. While there have been mixed reaction from techies and non-techies alike, the new technology included in the 4S is actually extremely impressive.

Ever been on a call on your iPhone and had it immediately drop at the first hint of a tree, overpass, or wall outcropping. We’ve all been there, and the less than stellar reception is merely magnified when you are talking to someone else who is also on an iPhone. Well the 4S has included two antennas within this model, one for transmitting and one for receiving calls. The camera is also new and includes improved shooting quality for both photos and videos.

Perhaps the most impressive improvement included in the 4S is the inclusion of natural language speech recognition technology called Siri. Siri operates at a level far beyond that of previous speech recognition engines, and in theory allows users to control the phone with their voice.

As anyone who has encountered interactive voice response (IVR) and speech recognition technology knows, most speech rec engines have a finite number of words and sounds that they are able to effectively recognize and dictate. Often times, these programs are not able to perform actions beyond a limited, extremely minimal amount of keywords, as the recognition technology is just not there.

With Siri, Apple has purportedly designed a system that allows for voice control of a phone by allowing it recognize natural language and considering both sentence structure and colloquial grammar. This is in part because of the Artificial Intelligence built in to Siri’s technology that allows it to recognize a vaster array of words, phrases, and sentences in order to properly execute user demands.

So what EXACTLY can Siri do? Send a text, find the weather, schedule a meeting, send an email, set an alarm, set a reminder, ask for directions, play a playlist, find a restaurant, and much more. The only thing it can’t do is make you a martini and tuck you in to bed.

This is a HUGE leap forward for speech recognition. To have artificial intelligence available that recognizes natural language and isn’t only operational through the use of limited vocabulary is revolutionary. If Siri works as well as it does in the Apple promos, this marks a huge milestone for speech recognition and voice technology on the whole.

Customers won’t start receiving their pre-ordered 4Ss until next Friday, but we will all have to do a bit more waiting to see if the phone and the technology delivers in the way it is promised. I for one can’t wait!

Comments are closed.