Apple's New Data Center Focus of Nuance Voice Recognition Partnership?

[JeffLebowski41]

Well for those of us that have build in integration in our cars, we feel a lot differently they you about good voice recognition...

agreed. I'm very much looking forward to this.

 

[KnightWRX]

The guy said "It's about what is done with those words. It's the natural language processing. The artificial intelligence. The computer trying to understand what it is you want to do and then doing it."

How is transliterating email to text and vice versa what he said?

Because doing Voice control in a data center has no logic ? Voice control is best done offline on the device itself. There is no logic in pushing out all the voice stream to a data center to process it and return a command to the device.

Hello useless data charges.

However, using a datacenter to convert voicemails to text and e-mails/SMS to voice makes sense. In the context of this thread, this makes much more sense, especially considering other platforms have this service in place (Cisco IP telephony Unity messaging or whatever it is called these days).

And my Sony Ericsson phone had Voice control back in 2002 and it worked fine. And my iPhone has it and has had it for the longest time, this would effectively be a non-story of Apple moving from one tech to another.

 

[johncarync]

Apple smarter than we think?

I'm starting to suspect that Apple began building this data center with no idea of what it would be used for. They just waited for the rumors to start flooding in and then they picked the best rumor and turned it into that. Sort of like Field of Dreams...build it and they will tell you what it should be.

 

[zoetmb]

To be fair, a major issue you didn't list (and probably the other guy's point) is being in public with other voices around. The voice software can't exactly control the microphone to block other voices. Not yet, anyway. And it will never be able to cope with noise that simply drowns out parts of your speech.

And yet there is music recognition software that can even recognize a song when there's an announcer mixed on top of the song and also recognize it via a very low-cost microphone (on the iPhone) and whether it's coming out of a 1/4" speaker with limited frequency response or a full-fledged system. I'm an ex-recording engineer and I understand waveform analysis, but I do not understand how this software (like SoundHound) is able to accomplish this.

So I think the state-of-the-art is advancing rapidly and I think they will get around the noise problem, both in hardware and in software. One of the ways you can get around noise at the microphone is to have two microphones, but you speak only into one. The microphones are wired out of phase to each other. Anything that goes into one microphone (your voice) passes through. Anything that goes into both microphones (the background noise) gets cancelled. This is much the same way that noise cancellation headsets work.

As for the utility of voice recognition, the main thing I would want it for is to command the iPhone when driving. To be able to say, "map current location" or "find new route home" or "call Steve Jobs" would be very useful. And as others have posted, a lot of speech recognition technology is already finding its way into cars.

 

[ipedro]

As for the utility of voice recognition, the main thing I would want it for is to command the iPhone when driving. To be able to say, "map current location" or "find new route home" or "call Steve Jobs" would be very useful. And as others have posted, a lot of speech recognition technology is already finding its way into cars.

No, more like: "Hey iPhone, can you show me where I am?" or "I want to take a different road home".

For Speech Recognition to gain mainstream adoption, it needs to understand plain English. Siri has already shown that it can. I think that it will be the defining feature of iOS5.

 

[charlituna]

Based on all the rumors, what DOESN'T the NC data center do?

Drink tiger blood, so ya know it ain't Winning!

Personally I think the rumors are pretty close to the truth in a way. I think there are a dozen rumors about what it is for simply because it was built for one thing, but for all of them. It is there to spread the load for all services and probably won't be the only one Apple builds. I can easily see them building another data center in Europe and one in Asia in the near future

It doesnt remove **** stains in your underwear. Thats the only thing I can think of.

No, but there is an app for that.

Yes - The thing I LEAST want in iOS 5 is voice recognition.

Tony

My brother is a physical therapist and his patients would disagree. They would love to have much better speech controls on their phones.

How long before TechCrunch reports that there are rumors that Osama bin Laden was hiding in Apple's NC data center?

(too soon?)

Well I did hear a joke about the US finding Osama by sending him a free iPhone 4

It doesn't run Macs...that's for sure. Ok, maybe 5% of the systems will be Macs.

Agreed. The actual servers are more likely unix/linux based with perhaps a couple dozen Macs for user terminals

 

[LightSpeed1]

Based on all the rumors, what DOESN'T the NC data center do?

lol

 

[DocNYz]

No, more like: "Hey iPhone, can you show me where I am?" or "I want to take a different road home".

For Speech Recognition to gain mainstream adoption, it needs to understand plain English. Siri has already shown that it can. I think that it will be the defining feature of iOS5.

Yeah for it to be a game changer it has to be completely effortless, or at least not force you to sit there and either guess what commands it will understand in what phrasing, or have you try and catch a glimpse at them floating by in the background of the screen.

I love the interface of Siri, but it's database isn't big enough and more importantly an improved, larger scale usage of that interface across all apps would be amazing. If you could use any service you want (like Zagat, or Facebook or IMDb) with quick voice integration, that would really change how you use the iOS devices. "What was that recent email Bob sent me about the new schedule?" or "What actress was in both Nightmare on Elm St. and the Social Network?" and having the phone parse those out and decide which data points to get from which databases and respond in a sentence or highlighted page will be much faster (and obviously more hands free if you're busy in a quiet environment) than unlocking your phone, deciding which app to use, scrolling through the names, comparing the differences on your own, etc.

 

[jwdsail]

The iPad 3 aka Knowledge Navigator 2 can't come soon 'nuf..

Think iPhone meets iPad meets Knowledge Navigator meets The Hitchhiker's Guide to the Galaxy.

Actually, I'd think advanced voice control/navigation in the iPad and MacOS X 10.7 Lion has much more potential than the iPhone.

 

[Loke2112]

Hey all I'm brand new here looking forward to helping and hopefully being helped along the way. I'm not sure if this has been posted as I am short on time to read all but the application Siri which uses this technology was bought I believe by Apple. I use it all the time to remind me to do stuff and it is way easier than typing it in. I think this had something to do with the Siri acquisition.

Tony

 

[Loke2112]

Yeah for it to be a game changer it has to be completely effortless, or at least not force you to sit there and either guess what commands it will understand in what phrasing, or have you try and catch a glimpse at them floating by in the background of the screen.

I love the interface of Siri, but it's database isn't big enough and more importantly an improved, larger scale usage of that interface across all apps would be amazing. If you could use any service you want (like Zagat, or Facebook or IMDb) with quick voice integration, that would really change how you use the iOS devices. "What was that recent email Bob sent me about the new schedule?" or "What actress was in both Nightmare on Elm St. and the Social Network?" and having the phone parse those out and decide which data points to get from which databases and respond in a sentence or highlighted page will be much faster (and obviously more hands free if you're busy in a quiet environment) than unlocking your phone, deciding which app to use, scrolling through the names, comparing the differences on your own, etc.

Just saw this. What he said...