Slashdot Mirror
Apple's Siri May Soon Process Voice Locally On a Device, No Cloud Required (appleinsider.com)
Proudrooster writes: "Apple wants Siri to become more useful to users when not connected to the internet, including the possibility of an offline mode that does not rely on a backend server to assist with voice recognition or performing the required task, one that would be entirely performed on the user's device," reports Apple Insider. Just give it 10 years and everything old is new again. Siri will join the ranks of Ford/Microsoft Sync and Intel Edison. Do any other phones/cars/speakers have this option right now? The new capabilities are outlined in a recently-published patent application that describes an "Offline personal assistant."
"Rather than connected to Apple's servers, the filing suggests the speech-to-text processing and validation could happen on the device itself," reports Apple Insider. "On hearing the user make a request, the device in question will be capable of determining the task via onboard natural language processing, working out if the requested task as it hears it is useful, then performing it. "
"Rather than connected to Apple's servers, the filing suggests the speech-to-text processing and validation could happen on the device itself," reports Apple Insider. "On hearing the user make a request, the device in question will be capable of determining the task via onboard natural language processing, working out if the requested task as it hears it is useful, then performing it. "
French toast, m'ladies
(tips toque seductively)
Welcome to 1998: https://en.wikipedia.org/wiki/Microsoft_Speech_API#SAPI_4
Golf clap.
While this is indeed the functionality that I want and have wanted for many years, I doubt that the local processing and storage space will permit decent cloud free processing. At least not without significantly restricting the vocabulary.
I'm not getting my hopes up.
Sorry, I didn't get that...
If so, how can they patent it? Used to be that patents would be made novel by tagging "in the cloud" to a mundane process, now the reverse is true, lol.
20 years from desktop to pocket sounds pretty impressive to me!!
And frankly it will probably work better than the old desktop stuff did which moistly did not take off (though Dragon seems to have done well with desktop software).
"There is more worth loving than we have strength to love." - Brian Jay Stanley
anything
but on a computer
but on a portable computer
but remotely on a server with a local display
but on a smartphone
but on a smart watch
but without internet
how can patents be granted for all of that ?
This really shouldn't be patentable. We had the ability to control computers with voice a quarter-century ago. Not only would there have been patents back then, but those patents would have expired long ago.
Do not look into laser with remaining eye.
... can perform some tasks offline. I can send an email, navigate, and change phone settings (adjust volume, etc), probably more. I've tested these in aeroplane mode and they worked fine.
Siri was a big disappointment almost from the start. Ask Siri any difficult question and all you get is search results. I mean I could have done that myself through a search engine. So any of these can tell you the weather, set a timer, do basic tasks, or be a sort of dumb assistant. But beyond the novelty its gets pretty old and useless after awhile.
What else is new?
Criminals are gonna be criminals. Even if they bought the laws.
... "I'm having trouble with the connection"
That drives me insane. Siri should at least handle voice calls when not connected.
The good news are hard to withstand, I don't know in what way I should answer them.
If this trend continues there will be a hard period of adapting.
The desktop computers doing it were 16MHz Macintosh LCII models with 4MB RAM
Let me check... nope still cool to have desktop software moved into mobile, even I the mobile processor is more advanced.
I have to think the stuff moving into mobile is a lot more accurate and can handle way more accents/languages than that ancient software though, so it's not like there has been no advancement that makes use of improved hardware.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
Because of course we all know that filing of a patent means the "invention" will hit the market real soon.
I discovered an open source project at https://snips.gitbook.io/docum... that already does this. This is not a turnkey device like Siri, but it seems to suggest the the cloud is not required to implement speech recognition.
sure it may soon evolve to be trained enough to process some queries without going 'to the cloud'; but apple is still going to get all the data, still going to get the voice samples, when you connect back up... because *that's* where the 'value' is in providing the feature in the first place. data. YOUR data. it's worth something to apple. enough to build an entire fucking ecosystem around.
It's much more efficient to upload the processed data to Apple's servers in order to keep tabs on all its users.
This reminds me of a place I have an account on where they occasionally get you to quickly take a photo in a certain pose to keep your account verified and it says *For privacy, your photo will not be uploaded! And then after you hit verify it quickly says "Uploading..." yeah, sure it's not uploaded.
half a year ago; reviewing how well ViaVoicehttps://www.youtube.com/watch?v=7CAkYs8PJT0 and Dragon Natural Speaking worked on vintage hardware back in the day: https://www.youtube.com/watch?...
over half a year ago; reviewing how well ViaVoice https://www.youtube.com/watch?... and Dragon Natural Speaking worked on vintage hardware back in the day: https://www.youtube.com/watch?...
Except of course when my iPhone is plugged in & CarPlay takes over and I'll get the message you mentioned because I'm in a less-than-perfect-area for data reception. Like literally anywhere on Whidbey Island, WA. It also replaces the radio's controls so commands like "tune to Sirius 22" don't work. Well they sort of do - Siri will tell me "please use your car's controls for the radio" and it means the buttons. It's infuriating.
Ford's onboard dogshit maps have a better voice engine for addresses and offline processing. Of course also unavailable when using CarPlay.
Getting a little sick and tired of Apple and Tesla getting treated with kid gloves when they "rediscover" technology that "uncool' manufacturers were using 10 years ago and getting the credit like they invented it.
My Garmin nüvi 3597LMTHD GPS has done this since I bought it in 2013. It's not connected to anything, not wifi, not cellular, and not bluetooth. Where I live (the boonies) the traffic features and even the map updates are pretty pointless, so there's never been a need to connect it to anything. Yet it understands me just fine. And unlike Alexa and others, it allowed me to rename it — it only responds to "yo, bitch", which is just how I like it. :)
Is the unit's understanding of language in general up to par with todays systems? No. But does it work for what it needs to understand? Yes. Very well.
For the home, when and if MyCroft gains a local speech understanding capability, that's the way I'm going. Everything I want to do is local, and the unit can be customized to run just about anything you put together (of course, commercial products aren't that easy to figure out, but that can be done in many cases as well.) Everything that depends on the "cloud" has failures, comm losses, and security concerns. Local is definitely the way to go.
Otherwise, everything you say ends up sent to Google, Amazon, Apple or whoever. And whoever they partner with / roll over for / get hacked by.
I trust Apple a little bit more as they've been pretty clear about being privacy focused, but that door is open for them to do "whatever" with your data, and it is best to keep that in mind. If they go local, that'd be nice. But inasmuch as it's a closed system, whereas MyCroft is an open system... yup, still going MyCroft if they can pull this off.
I've fallen off your lawn, and I can't get up.
I've always assumed that the real reason to send the sound for external processing is so that it can be stored and analyzed.
Smart phones have been powerful for a long time.
Maybe now they can just process the voice locally and then send the data to the collection center.
iOS already processes voice locally on the device. Cloud is only required for the Siri stuff. As proof, set an iOS device into flight mode, and open anything with an on-screen keyboard: edit a note, draft an email, etc. Tap the microphone icon, and talk. You'll see your speech transcribed with no resort to the cloud. (Misleading Headline is Misleading -- Film at 11.)
To your point, in notepad, the microphone works for text to speech, but voice commands won't play a local song on the phone or make a phone call.
Try saying:
Open iTunes
Play Playlist Chillout
After waiting 60 seconds:
"I'm having some trouble with the connection. Please try again in a moment."
So yes, it has speech to text, similar to the Commodore 64, but no commands work, even if the command doesn't need the Internet for anything such as playing a song loaded on the phone or opening the camera.
Sure -- cloud is required for the Siri stuff. And it makes sense that Apple would want to move Siri smarts for stuff that doesn't essentially require an internet connection (playing local music, etc) onto the device. My point is just that the Slashdot headline is misleading, conflating voice-to-text with Siri.
(And the C64 had (as you say) text to speech, but most definitely did not have speech to text, which is of course orders of magnitude harder.)
BS, cant even enable that microphone without agreeing to send shit to apple servers. Can’t test.
That's a feature that medics almost certainly don't know about that would be *super* useful to them. No need to get medical secretaries to transcribe dictated notes. And no data being processed in the cloud (= offshore and thus illegal, if you're in the EU). And the notes can be password protected too.
Great, the voice processing is done locally. You still need to be connected however to get any *data* back. Or does Siri also cache a local copy of the entire internet?
Just put it on the home screen. You can do that with contacts, and almost anything else. Even arbitrary Android "intents".
Tapping on icons three times, on a 3x3 grid, is not harder than saying three words, with a GUI without delays.
It is amazing, how many iDiots don't even realize, that it does nothing more, than stream your audio to their server, and receive a command back.