Apple Patents for Speech Recognition
Mac Rumours is reporting that Apple filed for <a href="http://www.macrumors.com"; target="_blank">"a number of microphone/speech recognition patents".</a>
Is this just regular Apple product development stuff that is filed away never to see the light of day or is this related to upcoming products? If so does this relate to speech recognition software in the entire Mac lineup? Or is this slated for the rumored update to the iPod? Or a competitior for the new Palm Tungsten T?
Personally I think it is a Sunday and Macrumors are desperate for a story
But I'd love to see speech recognition/voice recording capabilities in an iPod or handheld device from Apple...
Is this just regular Apple product development stuff that is filed away never to see the light of day or is this related to upcoming products? If so does this relate to speech recognition software in the entire Mac lineup? Or is this slated for the rumored update to the iPod? Or a competitior for the new Palm Tungsten T?
Personally I think it is a Sunday and Macrumors are desperate for a story
But I'd love to see speech recognition/voice recording capabilities in an iPod or handheld device from Apple...
Comments
<strong>Wrong Forum</strong><hr></blockquote>
Sincere apologies. Please move immediately to the GQ Fantasy Women Forum.
edit: oops, already moved, but not my initial choice for the change
[ 03-02-2003: Message edited by: Retrograde ]</p>
[quote] An improved speech recognition device is provided. The speech recognition device comprises a display with at least two built in microphones and a speech recognition module electrically connected to the display. The speech recognition module uses an algorithm that may take into account the position of the built in microphone on the display. The display may have a first axis of rotation where the microphones may be placed an equal distance from the first axis of rotation. <hr></blockquote>
I was under the impression that it would take Lot's of power and smart algorithms to increase accuracy. I look forward to any movement in this areas since Speech Recog is a Holy Grail of computing that would literally "Change Everything"
Or maybe we'll get Speech Extreme, where Apple ropes the poor video card into doing the work. <img src="graemlins/lol.gif" border="0" alt="[Laughing]" />
For me, the holy grail would be an iMac with motors on its neck, so that it could detect your position with its stereo mic's and turn to look at you when you call its name. The Pixar lamp will be a reality!
either this is just regular R&D stuff (most likly)
or some new killer app that is truly revolutionary
like a vioce prompt, a Q and A session with the computer insted of clicking your way through countless menues and options. (unlikly but cool)
photo shop human "new window"
computer "RGB or "CYMK"
human "RGB"
<strong>Maybe not right now, although there is AltiVec, but perhaps with the aid of the 970, and some of the onboard logic that moki has hinted at, it could happen.</strong><hr></blockquote>
Great. I can see it now. The day we get the 970, the Mac platform becomes twice as fast and then, by the time "The Steve Jobs Technology Hour" has ended, they're back to slow again. When will our Macs simply be friggin', flippin', undeniably, unbelievably fast? Apple, stop spending our CPU clocks before we have them!
<strong>or some new killer app that is truly revolutionary
like a vioce prompt, a Q and A session with the computer insted of clicking your way through countless menues and options. (unlikly but cool)</strong><hr></blockquote>
That was done like, ten years ago or something.. I've read about several programs that could do this, even though I do expect they werent that good, since their popularity arent that big....
Maybe we get Mac OS Xv sometime! (Xvoice)
Nevertheless, I'm interested to see what Apple has planned for us. Post-Steve, they don't do just-for-the-heck-of-it R&D.
probably apple R&D
what i'd like to see is the iPod with voice recognition commands
<strong>
what i'd like to see is the iPod with voice recognition commands</strong><hr></blockquote>
Yeah, like I'd really want to speak to my iPod. I honestly cannot see a use for this at all...where would it come into play in day to day life.
The only thing i could envisage being useful is if it recorded voice - but recognising it would be damn hard to do on a processor as feable as the one in the iPod.
[ 03-02-2003: Message edited by: low-fi ]</p>
<strong>
Yeah, like I'd really want to speak to my iPod. I honestly cannot see a use for this at all...where would it come into play in day to day life.
The only thing i could envisage being useful is if it recorded voice - but recognising it would be damn hard to do on a processor as feable as the one in the iPod.
[ 03-02-2003: Message edited by: low-fi ]</strong><hr></blockquote>
ur driving down the highway with ur iPod laying on the passengers seat...
"iPod(some command to let iPod listen) - skip" -goes to next song
"iPod - skip" -goes to next song
"iPod - shuffle play" -shuffle play enabled
--depending on HOW advanced it is--
"iPod - artist keller williams" -3secondn wait and a list of kw songs are made and start playing
--super mega advanced
"iPod - artist keller williams song Get On Up" -that songs plays
-------------------------------------------
now i understand the current iPod processor probably couldn't handle this, and i dont expect apple to make this, because like you said...well u said it in a derogetory fasion but i'll reword...there rn't really any super important needs for it, just an extra feature that if possible to add without driving up cost would be nice
I am thinking of an easy to set up, non-evasive home automation solution. Apple is already supporting bluetooth scripting applications for use with your cell phone. Why not place a bluetooth enabled mic in every room? "Lights on, lights off." That sort of thing. You could even do without the mic part (although now I am straying off topic). As long as you had some other bluetooth device in the room to acknowledge your cell phone's presence (and yours, of course) the lights would go on.
Of course, you would need a computer to electrical hardwiring solution in place.. maybe thats not as simple as I first thought.
Apple is and has always been the market leader in Asia and most countries with non-roman alphabets... stroke recognition in Asian languages is far easier than in English...
who says this voice recognition patent refers to English?
I remember an Expo in the mid 90's where Apple showed Voice Recognition that correctly parsed sentences from a Scotsman as well as Chinese-accented English (both without training the system for the particular voice).
This was a big deal at the time, since it implied heuristic syntax over phonemes as much as waveforms, and all other types of Vox Recognition require extended training of the system with your unique inflection patterns.
Even early military systems for pilots were finicky... driving to the airbase, the pilot loads his dictaphone with "flaps", "eject", etc... problem was, at 6.5Gs with incoming SAMs, the pilots vocal cords and panicked sphincter don't produce the words "eject" in anything near the relaxed voice he trained the system in that morning, and the jet ignores him.
Apple's mid 90's system seemed to solve this problem by correctly interpreting the meaning of words from speakers it had never heard before (unless the crowd was full of ringers).
Struggling to remember now... did Apple patent the voice-security authentication-by-waveform of multiple user login that came in OS 9?
hmmmm
even the fastest secretaries max out between 130 and 180 wpm typing... humans can reasonably speak close to 300 wpm... an "90-95% accurate" voice-recognition system would dramatically increase data entry productivity... and I'm always going to proof for spelling and grammar anyway, so I won't sweat the missing 5% perfection... but would dictation make life faster than typing in some environments... yep.
optional whisper mode, of course
</devils_advocate>
[ 03-03-2003: Message edited by: curiousuburb ]</p>
<strong><devils_advocate>
Apple is and has always been the market leader in Asia and most countries with non-roman alphabets... stroke recognition in Asian languages is far easier than in English...
who says this voice recognition patent refers to English?
I remember an Expo in the mid 90's where Apple showed Voice Recognition that correctly parsed sentences from a Scotsman as well as Chinese-accented English (both without training the system for the particular voice).
This was a big deal at the time, since it implied heuristic syntax over phonemes as much as waveforms, and all other types of Vox Recognition require extended training of the system with your unique inflection patterns.
Even early military systems for pilots were finicky... driving to the airbase, the pilot loads his dictaphone with "flaps", "eject", etc... problem was, at 6.5Gs with incoming SAMs, the pilots vocal cords and panicked sphincter don't produce the words "eject" in anything near the relaxed voice he trained the system in that morning, and the jet ignores him.
Apple's mid 90's system seemed to solve this problem by correctly interpreting the meaning of words from speakers it had never heard before (unless the crowd was full of ringers).
Struggling to remember now... did Apple patent the voice-security authentication-by-waveform of multiple user login that came in OS 9?
hmmmm
even the fastest secretaries max out between 130 and 180 wpm typing... humans can reasonably speak close to 300 wpm... an "90-95% accurate" voice-recognition system would dramatically increase data entry productivity... and I'm always going to proof for spelling and grammar anyway, so I won't sweat the missing 5% perfection... but would dictation make life faster than typing in some environments... yep.
optional whisper mode, of course
</devils_advocate>
[ 03-03-2003: Message edited by: curiousuburb ]</strong><hr></blockquote>
people speak like 150-200 wpm or something, not 300...u can interpret 300-400 - that what u thinking of?
<strong>now i understand the current iPod processor probably couldn't handle this, and i dont expect apple to make this, because like you said...well u said it in a derogetory fasion but i'll reword...there rn't really any super important needs for it, just an extra feature that if possible to add without driving up cost would be nice</strong><hr></blockquote>
Sorry ast3r3x - I didn't mean it like that!
Though i could use voice recognition in the car: now that would be useful
low-fi