An anonymous reader writes "Researchers at Dartmouth College have developed new software that uses the microphone on the iPhone to track and interpret a user's everyday activities using sound. The software, called SoundSense,
picks up sounds and tries to classify them into certain
categories. SoundSense can recognize completely unfamiliar
sounds and runs entirely on the phone. It automatically classifies
sounds as 'voice,' 'music,' or 'ambient noise.' If a sound is repeated
often enough or for long enough, SoundSense gives it a high 'sound
rank' and asks the user to confirm that it is a significant sound and
offers the option to label the sound. In testing, the SoundSense
software was able to correctly determine when the user was in a
particular coffee shop, walking outside, brushing her teeth, cycling,
and driving in the car. It also picked up the noise of an ATM and a fan in a particular room. The results [PDF]
of the experiments were recently presented at the MobiSys 2009 conference."
No kidding. Yet another way to track people's whereabouts. What is this, a way to close the gap when there's no GPS reception? Are we going to have to take the battery out of our phones when we're not actively using them? Perhaps install a hardware switch on the phone to disconnect the microphone? WTF?
But in contrast to previous technologies, at least this runs standalone on the device, rather than as a web service.
Also, it doesn't record everything verbatim, but rather just tries to find characteristics of different environments, and classify them. (I don't have direct knowledge of this, but it would be very resource-intensive and pointless to record all the ambient noise used to recognize you're in the office, for example).
I think it is a sensible idea. Obviously humans use their senses to be awa
GPS can generally tell you were you are, but only up to a point. Are you in the coffee shop, bathroom, or conference room in that giant building located at 455 N. Example?
Not only did you get it wrong, but most of the people replying to your post got it wrong as well.
GPS comes in several varieties. The common one used in cheap handheld units are using the C/A signal [wikipedia.org]. This gives you an accuracy down to about 5 meters (due to multipath and atmospheric interference). The accuracy is determined by the precision of the built in clock which in term determines the size of your unit. C/A gets you down to about 5 meters accuracy.
Add in DGPS [wikipedia.org] and you more or less eliminate atmospheric interference which can get you down to about half a meter. Technically 20 inches can count as a few inches when compared to the 200 otherwise.
P-code (the military) gives you down to about 2 meters accuracy by comparison. Not sure how much better they get with DGPS, but I'd suspect it'd get them to 1/10th just like the C/A does.
For the best accuracy you won't be relying on L1 and L2 directly (decoding the signals), but will be looking at the carrier phase change [wikipedia.org] which requires bigger and better antennas as well as a much more precise clock which is why most of these are big, bulky and used for surveying more than anything.
If you're moving around (airplanes pictures and very likely road surveying as well) you can get down to about half a centimeter, but expect from 5 mm to 10 cm). When not moving for a significant amount of time, you can get down to 2 mm to about 3 cm)
As for one of the replies claiming that GPS only gives you a 2D location, this is rubbish. You need a clear view of a minimum of four satellites to get a proper height as well as longitude and latitude. The accuracy of each of these coordinates varies as well . However the biggest inaccuracy you're likely to face when dealing with GPS is using the wrong datum [wikipedia.org] for your map. It's fairly easy to end up with coordinates several hundred meters from the correct one, merely by forgetting to switch datum when moving into a new area.
If you're not using the US GPS but instead rely on Glonass [wikipedia.org] (not done yet) you get some advantages. Since each satellite is running its own discrete frequency, you're essentially able to rule out atmospheric errors. This adds to the cost of the unit though, as it now needs a much more advanced radio receiver. I can't remember if Galileo [wikipedia.org] uses one or multiple frequencies. I think it's supposed to use two, but it's not that important.
Granted, it's been a while since I worked heavily with GPS theory, so feel free to correct me (if you can cite properly). I'm using Wikipedia because I doubt most of you are able to read Danish which is the language my text books on the subject.
the SoundSense software was able to correctly determine when the user was in a particular coffee shop, walking outside, brushing her teeth, cycling, and driving in the car
at least this runs standalone on the device, rather than as a web service
This sounds more like proof-of-concept than anything practical...for now. I'm not so worried about the "government" using this technology as I am concerned that it is a marketer's dream.
It's curious that there is so much concern about the government having some sort
so in other words this will effectively let me tweet when i'm pooping without having to type it up on my phone. it may also be handy finding out if people have washed their hands after such an event;)
Lets be honest people. If a device is capable, someone will write the software to enable it. This shouldn't be surprising or shocking. When 'wearable computers' started getting buzz it was because people were walking around with web-cams attached to their heads seeing everything they could see and slashdot thought it while amazingly geeky, was cool. This isn't that different except there's no soldering required.
To be honest, we haven't even seen the worst of it yet. Considering the deluge of FPGA and EEPROM powered embedded devices out there you'd best be scared of the things that are _hard_ to reprogram, not the ones with complete IDEs and API documentation available.
I'm more concerned about someone snooping on me from my Jura Capresso than I am from my cell phone.
> To be honest, we haven't even seen the worst of it yet. Considering the deluge of FPGA > and EEPROM powered embedded devices out there you'd best be scared of the things that > are _hard_ to reprogram, not the ones with complete IDEs and API documentation available.
The latter are the ones that someone can reprogram remotely without your knowledge.
I would posit that the 'hard' to program devices will be the ones that will go unnoticed for longer as there's no introspection or prosumer/developer tools available to identify aberrant behavior. If your DVR stutters occasionally when playing back it could be a bad MPEG-TS packet, or because the packet sniffer stomped the decoding threads priority.. How would you ever know?
Actually not a bad idea. A lot of time I would like to wear my iPod while doing house work, but I don't because I can't hear the baby cry. If I had an iPod that could recognize when the baby was crying, and play it over the headphones in place of my music, so that I knew the baby was crying, then I would really appreciate this feature. Same goes for somebody calling out my name. Even if there was a 1 second delay, it would be awesome.
Now That's some creative thinking. I've been trying to figure out a practical use for this technology other than the article's lame example of turning off the ringer when you're in a meeting. Recognizing sounds that might be important -- baby crying, doorbell, other phone, barking dog -- so you can listen to something else. Nice idea.
Then, if male: "Please, remember that there are ladies in the house, and put the seat down when you're done." Else, if female: "If you're out of toilet paper again, I can call your sister to bring it to you."
"The SoundSense software was able to correctly determine when the user was [...] brushing her teeth [...] Choudhury says that enabling the software to learn to recognize new sounds will be essential for practical applications. 'A system that can recognize sounds in a person's life can be used to search for others who have the same preferences'"
That sounds like great functionality. " If you like brushing your teeth, you may be interested to know that 21 out of 23 people on your Contacts list also enjoy brushing their teeth."
It would probably go more like :
Phone says : I hear that you're brushing your teeth.
Pause for connection. Crelm toothpaste with the miracle ingredient, Fraudulin! Choose Crelm toothpaste! To order Crelm toothpaste with the miracle ingredient, Fraudulin, press the "Buy" button now.
could be to anonymously feed thousands of opt-in users current sound input into a public algorithmic service which provides data facilitating the creation of truely random numbers. Similar to the use of atmospheric noise by sites like Random.org; but this could be better because the sound input devices are decentralized and always moving.
There are the obvious problems to overcome with attempts to game this sort of system, but I think it an interesting idea.
Or you could just use the quantum resistor noise from the input impedance of your sound card [digital-scurf.org]. Seriously, people seem to think getting good quality quantum entropy is hard. It isn't. The hard part is turning that entropy into bits: ensuring you have enough entropy to be generating as many bits as you are, and making sure that those bits are whitened and debiased properly, and that the entropy is properly distributed among them so that the non-random stuff in your input stream doesn't leak through. Not to
OK, so jokes aside, could not this be used to build profiles of areas where you might be trying to hold a coversation and apply noise cancellation activily to the phone? blocking out all the background noise for the person you are talking too, perhaps even making it a bit eaiser for you to hear in the process?
Finally, an app that can automatically record how many times a day we were rejected by women. It should be called iGiveUp and have an icon of a geek with a gun in his mouth.
The application in question categories sounds and thus tries to label them. This is typically done using AI algorithms like a combo of a neural net and fuzzy logic. That is pretty impressive to have it running on a mobile phone of all things.
I think it's very cool that we have come to the point that we basically have a device in our pocket that resembles a combination of the stuff mentioned in books like the Hitchhiker's Guide and the Ender saga.
And it rings you to ask if it's a significant sound?
After 30 minutes...I'll get a buzz saying:
"Do you want me to remember the fart sound I keep picking up? And by the way, you better get some fresh air"
... though I suppose most of the Slashdotters would say "Someone really is out to get me!" Maybe. Let's ignore that for a moment, and address a question that arose multiple times in the comments. What possible use was this developed for?
Is there a potential for abuse? Sure. There's lots of things with abuse potential. But I hate to break it to you, but they tracked Pablo Escobar in 93 from his phone. And it's only got easier since. If you want perfect privacy, don't get a cellphone.
Perfect privacy is impossible unless you never come into civilization. Detective work, for the longest time, has been about defeating privacy.
Example: You go AWOL because you didn't want to get sent to (insert next big US deployment) to die in the miserable conditions against an enemy that you don't have permission to shoot. You know they'll come and haul you to jail for many years, so you simply don't get a phone. The military calls your house, and of course, your roomie/parents/etc say that they haven'
People don't care about privacy, they just don't. Everyone here on/. upset about govenment civilian spy programs and such, and then everyone else gives up everything about themselves, willingly and deliberately, on sites such as Facebook.
Let me say that again, just so it's clear - most people just/don't care/ about privacy.
And/they/ are the ones that we need the laws to protect, those not smart or competent enough to take care of themselves.
Bringing the point back to the article - as soon as some flashy
I see a lot of tags/comments asking what this is useful for. There are a few uber-nerd things like recording your life and whatnot that I'm not going to get into, but the big one is determining location.
There are a TON of sweet things you can do with accurate location information, but the one that I'm most yearning for is to control my bluetooth, wifi, ringer volume, etc based on where I am during the day.
I'm an Android user, and there's a very nice applet called Locale that attempts to do this, but it proves to be pretty useless. The reason is that you're either using GPS (drains battery, doesn't work indoors) or wifi (drains lots of battery, and is the primary thing you want to control) to figure out where you are. If using the microphone and cpu is cheaper in energy, then this will be a big win.
Beyond the energy use argument, one of the main things you want to control is bluetooth - again, it drains batteries when on, and is not generally useful. But it's EXTRAORDINARILY useful in the car if you have a hands-free setup. Again, figuring out when you're in a car is hard via GPS or wifi, but this technique would seem to knock that one out of the park.
So, in summary, having your phone know where you are in your daily routine allows it to be more intelligent about what services and functionality it enables, and thus makes your cell phone that much smarter and more valuable.
There is no way in hell an app constantly running and listening to the mic is going to last 8 hours.
I'm on my second iPhone, both the original and the 3G, there is no way in hell they have the battery life to do this.
If by 'everyday life' the author means sitting in front of your PC with the USB cable attached, then okay, not my definition of everyday life, but okay. Otherwise, I call bullshit.
Arbitron is already using something somewhat like this.
They have their "test families", or whatever they're called, carry small devices they call "portable people meters [wikipedia.org]". Television programming includes sonic markers outside of the audible range, which these devices pick up via a small mic.
One would think that being able to identify television/radio programming without pre-inserting said inaudible watermarks could simplify their process.
Actually this sounds like the echo location batman used in the Dark Night at the end except for being self contained on the phone versus using just the mic from all phones and processing centrally
Privacy (Score:5, Funny)
For those of us who dont want to pay for it, dont worry! Next year the goverment will install it to your phone for free!
Re: (Score:2)
Re: (Score:3, Interesting)
Also, it doesn't record everything verbatim, but rather just tries to find characteristics of different environments, and classify them. (I don't have direct knowledge of this, but it would be very resource-intensive and pointless to record all the ambient noise used to recognize you're in the office, for example).
I think it is a sensible idea. Obviously humans use their senses to be awa
Re: (Score:3, Insightful)
GPS can generally tell you were you are, but only up to a point. Are you in the coffee shop, bathroom, or conference room in that giant building located at 455 N. Example?
Re: (Score:3, Informative)
GPS can tell you where you are within a few inches.
Your mapping software is what lacks in precision.
Re:Privacy (Score:5, Informative)
Not only did you get it wrong, but most of the people replying to your post got it wrong as well.
GPS comes in several varieties. The common one used in cheap handheld units are using the C/A signal [wikipedia.org]. This gives you an accuracy down to about 5 meters (due to multipath and atmospheric interference). The accuracy is determined by the precision of the built in clock which in term determines the size of your unit. C/A gets you down to about 5 meters accuracy.
Add in DGPS [wikipedia.org] and you more or less eliminate atmospheric interference which can get you down to about half a meter. Technically 20 inches can count as a few inches when compared to the 200 otherwise.
P-code (the military) gives you down to about 2 meters accuracy by comparison. Not sure how much better they get with DGPS, but I'd suspect it'd get them to 1/10th just like the C/A does.
For the best accuracy you won't be relying on L1 and L2 directly (decoding the signals), but will be looking at the carrier phase change [wikipedia.org] which requires bigger and better antennas as well as a much more precise clock which is why most of these are big, bulky and used for surveying more than anything.
If you're moving around (airplanes pictures and very likely road surveying as well) you can get down to about half a centimeter, but expect from 5 mm to 10 cm). When not moving for a significant amount of time, you can get down to 2 mm to about 3 cm)
As for one of the replies claiming that GPS only gives you a 2D location, this is rubbish. You need a clear view of a minimum of four satellites to get a proper height as well as longitude and latitude. The accuracy of each of these coordinates varies as well . However the biggest inaccuracy you're likely to face when dealing with GPS is using the wrong datum [wikipedia.org] for your map. It's fairly easy to end up with coordinates several hundred meters from the correct one, merely by forgetting to switch datum when moving into a new area.
If you're not using the US GPS but instead rely on Glonass [wikipedia.org] (not done yet) you get some advantages. Since each satellite is running its own discrete frequency, you're essentially able to rule out atmospheric errors. This adds to the cost of the unit though, as it now needs a much more advanced radio receiver. I can't remember if Galileo [wikipedia.org] uses one or multiple frequencies. I think it's supposed to use two, but it's not that important.
Granted, it's been a while since I worked heavily with GPS theory, so feel free to correct me (if you can cite properly). I'm using Wikipedia because I doubt most of you are able to read Danish which is the language my text books on the subject.
Parent
Re: (Score:3, Insightful)
This sounds more like proof-of-concept than anything practical...for now. I'm not so worried about the "government" using this technology as I am concerned that it is a marketer's dream.
It's curious that there is so much concern about the government having some sort
Re:Privacy (Score:4, Funny)
That was the sound of you shooting someone. For a monthly charge of $500 on your cell-phone bill, the phone won't call the police. Accept the charge?
Yes No
Parent
Re:Privacy (Score:5, Funny)
US consumers are safe for next 10 years though: even the government won't be able to force telcos to activate a new feature.
Parent
Re:Privacy (Score:4, Funny)
Why not?
Your bill can now have:
"Federal Government Location Audio Sensing Fee : $4.99"
on it.
Parent
new and improved ways to tweet your life (Score:5, Funny)
Re: (Score:2)
It's easy to say 'OMG PRIVACY!!' but.. (Score:3, Insightful)
Lets be honest people. If a device is capable, someone will write the software to enable it. This shouldn't be surprising or shocking. When 'wearable computers' started getting buzz it was because people were walking around with web-cams attached to their heads seeing everything they could see and slashdot thought it while amazingly geeky, was cool. This isn't that different except there's no soldering required.
To be honest, we haven't even seen the worst of it yet. Considering the deluge of FPGA and EEPROM powered embedded devices out there you'd best be scared of the things that are _hard_ to reprogram, not the ones with complete IDEs and API documentation available.
I'm more concerned about someone snooping on me from my Jura Capresso than I am from my cell phone.
Re: (Score:2)
> To be honest, we haven't even seen the worst of it yet. Considering the deluge of FPGA
> and EEPROM powered embedded devices out there you'd best be scared of the things that
> are _hard_ to reprogram, not the ones with complete IDEs and API documentation available.
The latter are the ones that someone can reprogram remotely without your knowledge.
Re: (Score:2)
I would posit that the 'hard' to program devices will be the ones that will go unnoticed for longer as there's no introspection or prosumer/developer tools available to identify aberrant behavior. If your DVR stutters occasionally when playing back it could be a bad MPEG-TS packet, or because the packet sniffer stomped the decoding threads priority.. How would you ever know?
Oh great.... (Score:2)
Like I really my phone to tell me my wife is bitching at me, or the kids are whining...
Re:Oh great.... (Score:5, Insightful)
Parent
Re: (Score:2)
Now That's some creative thinking. I've been trying to figure out a practical use for this technology other than the article's lame example of turning off the ringer when you're in a meeting. Recognizing sounds that might be important -- baby crying, doorbell, other phone, barking dog -- so you can listen to something else. Nice idea.
Re: (Score:2)
Just get some speakers!
Re: (Score:3, Interesting)
Re: (Score:3, Funny)
Like I really my phone to tell me my wife is bitching at me...
That's better than telling her you believe she's bitching at you!
Just point to the phone's labeling, smile and shrug helplessly.
ob. Clippy (Score:4, Funny)
"It sounds like you're going pee."
Then, if male: "Please, remember that there are ladies in the house, and put the seat down when you're done."
Else, if female: "If you're out of toilet paper again, I can call your sister to bring it to you."
Social networking? (Score:5, Funny)
From TFA:
"The SoundSense software was able to correctly determine when the user was [...] brushing her teeth [...] Choudhury says that enabling the software to learn to recognize new sounds will be essential for practical applications. 'A system that can recognize sounds in a person's life can be used to search for others who have the same preferences'"
That sounds like great functionality. " If you like brushing your teeth, you may be interested to know that 21 out of 23 people on your Contacts list also enjoy brushing their teeth."
Re: (Score:2)
It would probably go more like :
Phone says : I hear that you're brushing your teeth.
Pause for connection.
Crelm toothpaste with the miracle ingredient, Fraudulin! Choose Crelm toothpaste!
To order Crelm toothpaste with the miracle ingredient, Fraudulin, press the "Buy" button now.
Another use for this technology (Score:3, Interesting)
There are the obvious problems to overcome with attempts to game this sort of system, but I think it an interesting idea.
Re: (Score:2)
Re: (Score:3, Interesting)
Or you could just use the quantum resistor noise from the input impedance of your sound card [digital-scurf.org]. Seriously, people seem to think getting good quality quantum entropy is hard. It isn't. The hard part is turning that entropy into bits: ensuring you have enough entropy to be generating as many bits as you are, and making sure that those bits are whitened and debiased properly, and that the entropy is properly distributed among them so that the non-random stuff in your input stream doesn't leak through. Not to
Sounds cool (no pun intended) but (Score:2)
Now my phone will ask why it's 4:20 all the time ;)
Practical Application (Score:2)
Just What Slashdotters Needed (Score:2, Funny)
Artificial intelligence algorithms on an iPhone (Score:2)
The application in question categories sounds and thus tries
to label them. This is typically done using AI algorithms like
a combo of a neural net and fuzzy logic. That is pretty impressive to have it running on a mobile
phone of all things.
I think it's very cool
that we have come to the point that we basically have a device
in our pocket that resembles a combination of the stuff mentioned
in books like the Hitchhiker's Guide and the Ender saga.
How Long Before... (Score:2)
Re: (Score:2)
reboot (Score:2)
I wonder how many of these SoundSense users will end up having the "awooaaaa" sound of a booting Mac get classified as a "significant sound".
Really? Sounds? (Score:2)
And it rings you to ask if it's a significant sound?
After 30 minutes...I'll get a buzz saying:
"Do you want me to remember the fart sound I keep picking up? And by the way, you better get some fresh air"
Ah, the paranoia... (Score:3, Insightful)
Research. That's right. Research. Actually finding out what people are doing/eating/etc. is actually really hard. People's recollection is full of holes. Think about it: What were you doing exactly 1 hour ago? Most people will be able to name the task, but not what specific activity in that task they were doing. Where you talking? Where you typing. Where you taking a two minute mental break and staring off into space? Yeah. It's hard to recall these inane details. But things like how many people you have contact with, how often you're in public-public (and not in an office, etc.) are hard to get at. How many times have I been to the café down the street in the last week? Hell if I know. But if I volunteer for a study - let's say, time spent in a restaurant by number of colds people get (totally made up on the spot for the purpose of illustration), this way the researcher doesn't have to trust a potentially erroneous recollection... they can get another, good estimate. It's made to be opt-in, and for research of this, that or the other thing.
Is there a potential for abuse? Sure. There's lots of things with abuse potential. But I hate to break it to you, but they tracked Pablo Escobar in 93 from his phone. And it's only got easier since. If you want perfect privacy, don't get a cellphone.
Re: (Score:2)
Perfect privacy is impossible unless you never come into civilization. Detective work, for the longest time, has been about defeating privacy.
Example: You go AWOL because you didn't want to get sent to (insert next big US deployment) to die in the miserable conditions against an enemy that you don't have permission to shoot. You know they'll come and haul you to jail for many years, so you simply don't get a phone. The military calls your house, and of course, your roomie/parents/etc say that they haven'
how much battery power will this suck down? (Score:3, Insightful)
how much battery power will this suck down?
People don't care about privacy. (Score:2)
People don't care about privacy, they just don't. Everyone here on /. upset about govenment civilian spy programs and such, and then everyone else gives up everything about themselves, willingly and deliberately, on sites such as Facebook.
Let me say that again, just so it's clear - most people just /don't care/ about privacy.
And /they/ are the ones that we need the laws to protect, those not smart or competent enough to take care of themselves.
Bringing the point back to the article - as soon as some flashy
In response to "Why?" (Score:4, Insightful)
I see a lot of tags/comments asking what this is useful for. There are a few uber-nerd things like recording your life and whatnot that I'm not going to get into, but the big one is determining location.
There are a TON of sweet things you can do with accurate location information, but the one that I'm most yearning for is to control my bluetooth, wifi, ringer volume, etc based on where I am during the day.
I'm an Android user, and there's a very nice applet called Locale that attempts to do this, but it proves to be pretty useless. The reason is that you're either using GPS (drains battery, doesn't work indoors) or wifi (drains lots of battery, and is the primary thing you want to control) to figure out where you are. If using the microphone and cpu is cheaper in energy, then this will be a big win.
Beyond the energy use argument, one of the main things you want to control is bluetooth - again, it drains batteries when on, and is not generally useful. But it's EXTRAORDINARILY useful in the car if you have a hands-free setup. Again, figuring out when you're in a car is hard via GPS or wifi, but this technique would seem to knock that one out of the park.
So, in summary, having your phone know where you are in your daily routine allows it to be more intelligent about what services and functionality it enables, and thus makes your cell phone that much smarter and more valuable.
Real motivation behind this (Score:3, Insightful)
Bullshit (Score:2)
There is no way in hell an app constantly running and listening to the mic is going to last 8 hours.
I'm on my second iPhone, both the original and the 3G, there is no way in hell they have the battery life to do this.
If by 'everyday life' the author means sitting in front of your PC with the USB cable attached, then okay, not my definition of everyday life, but okay. Otherwise, I call bullshit.
Re: (Score:2)
Clippy! (Score:4, Funny)
It sounds like you are having trouble reaching orgasm. Would you like me to
1) Moan real loud
2) Show erotic video
3) or, just vibrate
Arbitron would love this (Score:3, Interesting)
Arbitron is already using something somewhat like this.
They have their "test families", or whatever they're called, carry small devices they call "portable people meters [wikipedia.org]". Television programming includes sonic markers outside of the audible range, which these devices pick up via a small mic.
One would think that being able to identify television/radio programming without pre-inserting said inaudible watermarks could simplify their process.
Beano Ads? (Score:3, Funny)
So if you fart a lot you'd get beano ads on your iPhone?
Re: (Score:3, Funny)
The EPA bills you for climate change.
Re: (Score:2)
Maybe it's missing a comma:
"...noise of an ATM machine and a fan in a particular room."
becomes:
"...noise of an ATM, machine, and a fan in a particular room."
or, more clearly:
"...noise of an ATM, machine (unspecified but noisy), and a fan in a particular room."
Re: (Score:2)
Or maybe they refer to a machine that makes ATM:s
Re: (Score:3, Interesting)