Slashdot Mirror
Voice Activated MP3 player
g0dsp33d writes "A US company is working on a voice controlled MP3 player for applications like cars where touch control is not as feasible. Considering technology like radar breaking and AI steering for robots, it reminds me of the possibility for a real life version of the car from Night Rider, KITT. Minus the cool jump effects, of course."
"Stop, what's that sound." It'll shut off right after it starts.
I'll actually be avoiding this, as I don't want anyone else to hear that I'm listening to Britney Spears.
Wait.
Damn it!
...and ITT was in Night Rider.
How much training is required to operate this thing reliably? With voice training, you can get any piece of software to ignore non-operator commands.
But I'm loath to sit out in the cold just to program the stupid thing in my sleigh.
How big a deal is it to take a fraction of a second to change the song anyway? We do it all the time with the radio, A/C, and speedometer already.
I'm sitting in a crowded place, listening to the White Album on my new voice activated MP3 player.
Me (quietly): Play Everybody's Got Something to Hide Except Me and My Monkey.
MP3 player: Please speak louder.
Me: Everybody's Got Something to Hide Except Me and My Monkey.
MP3 player: Please speak louder.
Me (shouting): Everybody's Got Something to Hide Except Me and My Monkey!!!
Everyone turns and looks at me. I make a mad dash for the exit.
Unknown host pong.
I dont find using my radio while driving to really be that difficult. I don't even have to look at it to change CD, tracks or alter the volume.
It seems to me that if the system was not perfect, IE it has common errors and played the wrong songs, that would make me more annoyed than the prospect of having to reach down to change something.
I guess you can call me crazy, but I still like adjusting dials and getting some feedback when I change the volume or change tracks, and i'm not sure I would like a voice activated system.
Can the Mp3 player turn itself off by making it play a soundtrack saying, "Shut down"?
;) No, it likely can't. Why? Because voice recognition absolutely sucks for noisy environments - ie, a car where there's music playing.
You're not the first smart alec to ask this.
It probably won't be able to understand you, let alone the music...
It's my biggest gripe with the iPod that nobodys built a voice activated accessory yet. When I'm drunk off my rocker and want to select a particular song, it takes forever. Overscroll, scroll back, overscroll, scroll back, etc. Very frustration. (I suppose it would also be useful when you've been running or just naturally get sweaty hands, as it's near impossible to use the touch wheel with moist fingers).
Okay, so an interface like that with the iPod is probably not possible, how hard could it be to introduce it on new iPods? Just say a artist, song name or album and get zipped right to it!
Hey now, let's not write off those cool Jump Effects too soon, okay?
it reminds me of the possibility for a real life version of the car from Night Rider, KITT
It's Knight Rider, and KITT stands for Knight Industries Two Thousand.
I don't want my car to talk to me. I want it to run on water
Or at least stop breaking and get 100 mpg.
Forget the talking.
Take the cheese to sickbay, the doctor should see it as soon as possible - B'Elanna Torres, "Learning Curve"
This kind of tech is a little too CPU intensive for a normal Digital Audio Player CPU (some of them can't even keep up with Vorbis) so I don't think there's much of a chance this will make it into any kind of portable player, which is a pity. There are plenty of situations (at the gym, biking, canoeing) in which this could be useful.. provided it works, of course.
Not a big loss, I guess, voice recognition software has always been a little sketchy (although I'm sure it's improved since my first experiences with it) and I doubt this will be all that usable in a convertable or even a sedan with numerous open windows. Nice idea, but I don't see it being any kind of revolutionary success..
Now the kiddies will be able to fight what music to listen to in the car. This will surely divert their parents' attention from driving. This is just like hooking up a clapper ("clap-on, clap-off" - a device that turns on/off lights when it detects a clap) to the lights of an auditorium full of people during a concert.
-Palal
They have a voice activated car with one of these in it....
;)
Telling it to "Back up" while you're cruising at highway speeds would end up with your transmission in the middle of the road and you looking at the bottom of a ditch!!
No unauthorized use. Trespassers will be shot. Survivors will be shot again.
I just don't think this is such a good idea. It would be great for cell phones (dial such and such), but for MP3 Players, I don't think that the technology is there, nor do I believe that anyone would want to use it. Imagine trying to play a song called "Sk8r Boi" or something similiar...
On top of that, why not just use steering wheel mounted paddle controllers? They keep your hands on the wheel and eyes on the road, and are very easy to access..
"Victory means exit strategy, and it's important for the President to explain to us what the exit strategy is." G.W.Bush
See subject.
-Kitt, find me some women!
-I'm sorry, I can't do that Dave.
will it suffer from clapper syndrom, like when you have a stereo plugged into the clapper and the song includes clapping?
idunno what the control words would be, but i bet there's enough songs with the word "louder" in it to make it suck.
A few thoughts:
How will this react to similar words to its commands being used?
if the song has the command words?
if the music is too loud for the device to hear the words?
if at least two people in a small area have the device and are trying to use it at the same time?
Does it only respond to a given vocal pattern?
Voice imprint?
Words that sound too similar to its commands?
Its command words not directed at itself (as in telling someone to stop doing something)?
Are there a bunch of armless drivers I'm not aware of? If you're too incompetent to take your hand off the wheel and adjust your music track, you probably shouldn't be driving anyway. Or allowed to own a car.
We are one consciousness experiencing itself subjectively. Back to you with the weather, Bob!
it reminds me of the possibility for a real life version of the car from Night Rider, KITT
When you have this. Floating of the Wings of Tenderness!
Comparing it to Windows will be a moot point, since El Dorado is going to have a 40% larger code base than XP.
Because voice recognition absolutely sucks for noisy environments - ie, a car where there's music playing.
I thought that the noise cancellation would be easier seeing that they have the source of the music, so they can phase it out 180 degrees.
But, other than that - voice recognition in a car would still suck (like it does everywhere else)
You can't expect to wield supreme executive power, just because some watery tart threw a sword at you
My cell phone can dial things now.
My guess is that the player isn't going to sort your music for you, you will need to put it in the genre you want.
Unless it's something really mind blowing, like it tells me what i want to hear next, and is actually correct, i am unimpressed.
http://www.mp3car.com
A lot of us hobbyists have done a lot of research and put a lot of computers in cars, with fabrication, touchscreens, DC-DC power supplies, and more.
I'm suprised Apple didn't do this...
or Griffin, or some other accessory company.
Michael: Faster, Knightboat! We gotta catch those starfish poachers.
Knightboat: You don't have to yell, Michael, I'm all around you.
Computer? Computer?
Ah yes, the mouse. [Holds up mouse to mouth]
Hello computer?
...it was Knight Rider.
You just made me feel incredibly old. Damn you, you insensitive clod!
But back on topic....
Considering how complex the car of the Future!® is going to be, with navigation systems and such, keeping the driver's eyes on the road is going to be a priority. The less a driver has to futz around with his eyes off the road, even for a moment, and one hand off the steering wheel, the better.
My question is how one goes about filtering the user's voice through the sound of the music. Would it be possible to make an MP3 that says basicallly, "I'm an voice-activated MP3 player virus. PLAYER TURN OFF!" *click*
"I am an Adept of Tantric VAX."
Hrmm couldn't this be used somewhat similar to a turing machine, except with audio cues instead of shapes? If you can voice controll playlist orders, and you have mp3's with commands such as "add moveup3.mp3 to bottom of list remove moveup.mp3"(assuming this command will remove first instance of moveup.mp3) moveup.mp3 of course is a audio sequence that moves the last song up one position in the playlist. With a complex sequence of such commands, couldn't you implement a turing-like rules system in audio?
Little Brother, watching the watchers
What if the song i'm listening to happens to have a voice command? Wouldn't that ruin my listening experience?
What changed under Obama? Nothing Good
Microsoft Voice Command provides a similar service for pocket pcs, through its integration with windows media player (for the pocket pc). I played with it, and it was novel for about a week, and then I realized that it was really not that difficult just to push one of the hard buttons, and I haven't used it much since. Though I must say I did attract some strange looks when I was shouting at my PPC.
In the beginning the universe was created. This made a lot of people very angry and is widely considered as a bad move.
Mac OS X already has voice recognition. Use a bit of Applescript, and with iTunes, you're off to the races..
-jcr
The only title of honor that a tyrant can grant is "Enemy of the State."
Reminds me of a Bash quote...
i got new car radio the other day, its pretty cool. you shout soul and it plays soul, you shout rock and it plays rock. the other day some kids ran past my car and i yelled "FUCKING KIDS" and it played michael jackson.
Forget the voice activated MP3 player. Why not opt for voice activated hydrolics? With something like that, girls would be (as Strongbad would say) "all up ons".
There is no facet of life couldn't be enhanced with a talking car.
Si tacuisses philosophus mansisses. If you had kept quiet, you would have remained a philosopher.
I'd rather have everything displayed within an integrated HUD that is projected onto the windshield (I know this has been done just can't remember the make and model of the car) and then add eye tracking components. Want to go to the next track? Glance at the next track button on the HUD. That's a lot better than trying to get voice recognition software to operate in a very noisy environment (at least the noise would be the music itself and at worse include external road noises).
And what if the music had vocal commands that could feedback into the system? Some song somewhere contains words such as 'play', 'stop', etc...
Shh.
iPod is better and iTunes is so great. I am completely happy with exactly what the iPod provides and I do not want any other options or features because it will cause the batter to wear down faster. The iPod just works and Apple design is the best.
It is about style, not just about listening to music on a portable player. The scroll wheel is awesome, also the iPod battery lasts longer then you think and it can be replaced cheaply.
Si tacuisses philosophus mansisses. If you had kept quiet, you would have remained a philosopher.
I don't know just how sophisticated technology in these things would be, but theoretically it shouldn't be too difficult to keep an audio system from responding to commands issued by the audio file. Voice recognition is done by analyzing the signal from audio, a process known as digital signal processing. When the signal matches something the system knows to look for, the programmed response is called.
:) My DSP experience is limited to working with a group of DSP people, and my project had nothing to do with theirs. I just picked up a fact here and there. :) Still, this seems a possible solution to me.
A major problem with voice recognition technology is when the technology cannot differentiate the speaker from the background noise. Not knowing what is reliable and what is noise, the system can be more prone to error. In this case, however, the system would know what it is outputting and to my thinking should be able to ignore any signal currently being output. Any incoming audio signal would be compared against the outgoing signal and if it matched it could simply be ignored. This should make it possible to eliminate noise coming from various sources, perhaps even something to detect noise outside the car and filter it out as well.
This is just my thinking.
I love my sig.
Awesome. I'm gonna get mine and I'll be like, "Tea, Earl Gray, hot." And then it will be like, "I can't do that, Dave," and then electrocute me.
"OH SHIT, THERE'S A HORSE IN THE HOSPITAL!"
Is that radar breaking, or radar braking?
Be seeing you.
scott
Just put the new Bose suspension system on the car if you want it to jump. Check out the last paragraph on page four of this article for a description.
what sig?
Eclipse car audio came out with a voice activated device a few years ago. It was called the Eclipse Commander, and controlled among other things: radio, cd player/changer, and power windows.
A simple example should suffice: "If you don't stop misspelling 'braking', I'll be breaking your nose."
I bought my CF-based MXP 100 from e.digital http://www.edig.com/techsupport.php#mxp over two years ago, and it's a voice-recognition player. Of course, to have it add that meta data, you must use their software to U/L the files into the device. And with a USB 1.1 connection, that's just not worth my time. Instead, I use a CF card reader & just write straight to the card.
What you want is irrelevant; what you've chosen is at hand! - Spock, ST VI
Basically, you'd have a folder buried in your System Folder called Speakable Items, and anything you put it in would be recognized when you spoke it. Activation was just as if you double clicked the file.
For example, there was a file in there called "Close this window" which was just an applescript telling the frontmost application to close the active window. There was also other stuff like little scripts which involve you in an interactive knock knock joke and things like that.
Anyway, being on a mac, none of my mp3's had file extensions and were just named things like "Walk On By", "Love Child" etc. I just put an alias (shortcut) for every mp3 I had into the Speakable Items folder, which made it easy to say across the room to the machine: "Computer, Love Child" and it would just start playing it.
Of course once something was playing it was impossible to stop or change tracks due to the music interfering with the speech recognition...
"Your door is ajar."
"Your taste in music is banal."
"RIAA validation code not found."
"Your door is closed and locked."
"Ingition disabled."
"RIAA lawyers and approprite law enforcement has been notifed."
"Please stand-by."
Awesome! I can just imagine the look on the cops' faces when their radar just cracks in half before they can get a reading.
Karma: Chameleon (Mostly affected by the 1980s)
I work in a loud environment where I use my MP3/WMA playing PDA to keep me from going insane.
Because I have to work with my hands and move around a lot, the player goes in my back pocket to keep the headphone wires out of the way. This works great until I need to adjust the volume or change tracks, which requires me to stop working, pull out the PDA and fiddle with it, and stuff it back in my pocket.
I would *kill* to have a way of telling the PDA what I want it to do either by voice command (but again, it's a loud workplace) or by telepathic command. A simple "next track" or "volume up" would do.
Somebody please invent this! Apple! Somebody?! Come on!
I really should be helpful instead of just being an ass...
braking
breaking
Karma: Chameleon (Mostly affected by the 1980s)
In the car, jogging, etc...
Combined with a bit of fuzzy logic, this might be a great addition. Instead of hitting a bunch of little buttons, one could just say:
"Rush shuffle volume 9 play" and get your named tracks played in shuffle order at a high volume with very little effort compared to all the little buttons necessary to get this done.
Of course, this will probably consume enough CPU power to sharply reduce battery life... Some bozo will add a damn button to turn on the voice function to compensate. The wonders of technology.
If these things all could be crammed into an Mp3 player, then lots of other devices could benefit. The mixture of voice and technology seems kind of like the laser was. Everybody knows it's cool, but we just haven't perfected the tech enough to catch up with the idea in practical ways --yet.
Blogging because I can...
AFAIK nasa experimented with this for operating stuff while under high g loading. Worked fine, except that people glance at things all the time. You could set it so that you need to stare at the icon for a time to trigger it, but I don't fancy that while driving.
They whose government reduces their essential liberties for temporary security, receive neither liberty nor security.
"Dead Kennedys - I Kill Children"
Angry looks from the surrounding commuters... Oh dear...
"Butthole Surfers - Kuntz"
Oh dear oh dear oh dear..
"Placebo - Evil Dildo"...
I think I'm going to get thumped! Seriously, I've had enough of morons and their 100dB(sic) headphone leakage let alone them shouting instructions into their iThingies over the traffic, trains, crowds.... I thought people wore and used these things to gain a little "peace".
Time flies like an arrow. Fruit flies like a banana.
When you're drunk off of your rocker, the mp3 player will just explode trying to figure out what your drunk ass is saying to it.
..that one song..geez..." *splat*
"Uhhh plesh play me theeee.. uhhhhhh... y'now..*hic*
Why not?
Actually, longer words and sentences are easier for computers to recognize because they differ more than short commands.
I've worked for a long time with this stuff.
There is a huge difference between matching one name from a set of pre-recorded by the same speaker (normally not more than 10), as the case is in the phone, and recognizing more complex sentences, natural languages, from any speaker.
:-)
We developed several years ago a natural language MP3-player, where, using ID3-tags, you could say things like "play some romantic music" (matching on the genre), or "enqueue songs by Abba".
Lots of hidden issues, from voice user interface design, natural language parsing, making it multilingual etc, and having it run on embedded hardware. We got quite far though.
If you're interested, check out our company at www.voxi.com. Our VCs went bust, so there's not much going on at the moment, but we have some nice IP if anyone's interested
You meant Knight Rider with the K!
Ant(Dude) @ Quality Foraged Links (AQFL.net) & The Ant Farm (antfarm.ma.cx / antfarm.home.dhs.org).
If it is voice activated, then songs that contain:
"next" "forward" "back" "stop" "go"
and are played on the in car system would have to be filtered out (i.e. check what was just played, and cancel that from the microphone) this will also allow you to whisper to it even when playing aqua at volume 11.
Except you might actually die.
#hostfile 0.0.0.0 primidi.com 0.0.0.0 www.primidi.com 0.0.0.0 radio.weblogs.com
The player would be vulnerable to a mp3-based DoS-attack: Kelly Osbourne shouts: "Shut up !" and the player shuts down...
Wonder if we can get it to do proper fuzzy searches of song titles.
"Play that one, you know, the one with the miller..no,not that one, the other one, you know which one I mean, the dulux adverts."
"How fine you look when dressed in rage."
This reads more like a press release of an "amazing new product" that they haven't actually built yet than some concrete information. If mobile phones have voice control, and mobile phones can play MP3s, how is "a voice controlled MP3 player" suddenly an amazing new advance?
Of course you could just go out and buy one now if you actually wanted one for your car. There's even a funky flash demo.
Meep meep
Mercedes-Benz (or rather, Daimler-Chrysler), for example, has been shipping an electronics system, LINGUATRONIC COMAND (for Cockpit Management and Navigation), for at least three years that is voice controlled. Voice recognition controls the radio, the CD, the integrated Motorola telephone. A 30-word vocabulary doesn't sound like much, but it gets the job done.
It's activated by a steering wheel stalk and is somewhat modal (but an MP3 player wouldn't have as many modes in the first place), but the hard part is all done by voice recognition.
I'm pretty sure other car manufacturers are shipping similar systems by now.
"It is our blasphemy which has made us great, and will sustain us, and which the gods secretly admire in us." - Zelazny
i've had an edigital mp3 player with voice navigation for four years now. http://www.edigital.com/product.php
Do you guys need a /sarcasm tag? Mod this up because it's hilarious.
Ill be excited when I can use commands like "play that one song, where it goes really high, and then makes that boom boom boom noise", or maybe if you could just wistle a couple of bars and it could figure out what you wanted to hear. That would be nifty, becase I am lazy and having to remember names of songs is annoying.
Please refer to these as Music Players, not MP3 players since they usually support multiple formats.
:T:R:A:N:S:
Reminds me of the joke that goes something like this:
.... and Michael Jackson comes on.
A lady goes to a car dealer to get a fancy car. The salesman shows her a model with a voice activated radio. He says "classical", and a classical station comes on. He says "rock" and a rock station comes on. Impressed, the lady decides to take it for a test drive. As she is pulling out of the lot, she narrowly avoids hitting some children and she yells "fucking children"
Have you metaroderated recently?
Are you serious? Do you look at your foot to find the brake pedal when those pesky kids jump out? Do you look at the indicator stalk when you signal?
They whose government reduces their essential liberties for temporary security, receive neither liberty nor security.
Doesn't work properly. More frustrating and distracting than pressing a button on the steering wheel.
I will also avoid this thing, but for different reasons.
Starting virus_song.mp3...
MP3: Hello, I sending you those in order to have youre advice my dear friend!!!!
Me: What?
MP3: MP3 player, volume up!
Me: No, volume down!
MP3: MP3 player, volume up! volume up! volume up!
Me: Down, damn it! Down!
MP3: up! up! up! up! up! up! up! up!
Me: MP3 player, please--
MP3: Lalalala! I can't hear you! Lalalalalalalala!!!
Me: *sigh*
MP3: MP3 player, volume max! Car, open windows! Attention! To all MP3 players in range: download virus_song.mp3 and start playing!!!!!
From other cars: Hello! Hello.. Hello!!! I sending you-- *crash*
How long before we read about the first accident caused by this new technology?
Sincerely,
Pan Tarhei Hosé, PhD.
"Homo sum et cogito ergo odi profanum vulgus et libido."
Although it would be handy to have a radar controlled system to brake the car to slow it down if it detected a potential collision.
Am I being nitpicky? I don't think so - keep in mind this wasn't posted by some random
How did Sepultura make that list? They are NOTHING like the other bands you named, and 98% of the Slashdot crowd has probably never heard of them.
That's just random.
p.s. I was a big Sepultura fan in high school back in 90-94.
http://www.epinions.com/elec-review-6E7E-2859892-3 9CD5C1D-prod1
and uses technology from Sensory http://www.sensoryinc.com/html/showcase/showcase.h tml#8.
I guess the novelty depends on how far they're taking the voice-recognition.
I'd guess they're going for pattern matching of a static number of commands, it's been done before (with varying success) - hey even mobiles do it today.
If on the other hand they're going to provide speech to text conversion ("Play unforgettable fire"), that's an entirely different (and processing-intensive) thing.
Somehow I have this image of the radar wave smashing into the car in front, crushing it into tiny little bits... - perhaps you meant braking, as in to brake, as in to decelerate?
AAAAAARGHGHGHGH! I feel better now.
I got my MXP100 off ebay for 40USD
It takes CF and accepts voice commands.
www.edigital.com
That would be trivial. But if it could switch itself on by playing a track saying "switch on", I'd be impressed.
Confucius say, "Find worm in apple - bad. Find half a worm - worse."
First ever car radio virus for the voice activated system
Disguised as the newest boy-band album, it features such tracks as:
"shuffle on, shuffle off, shuffle on, shuffle off"
"volume up, volume up, volume up, volume up"
"mute"
and my personal favorite, "format memory card"
"Stereo"
"Stereo Ready"
"Playlist 'Mellow'"
"Activating playlist 'Mello'"
time passess....
"Stereo"
"Stereo Ready"
Car comes out of nowhere from the side, crossing just in front of driver and goes into a tree on the side of the road - accident avoided by inches.
"SHIT!!!!!"
"Activating playlist 'Boy Bands'. No need to yell."
"There is more worth loving than we have strength to love." - Brian Jay Stanley
I'm not a spelling nazi unless it appears on the FRONT PAGE.
Gamingmuseum.com: Give your 3D accelerator a rest.
I wondered what ever happened to John Oates.
free online diet tracking.
What's much simpler to implement, possibly more practical, and almost as nice is speaking mp3 players so you can operate it without looking at the screen. Rockbox firmware supports this to some extent.
Confucius say, "Find worm in apple - bad. Find half a worm - worse."
It takes maybe two seconds to say "Stairway to Heaven". On my iPod, it takes up to six button presses and a bunch of scrolling to seek to Stairwy to Heaven.. During that time, my eyes are off the road (which is why I usually program playlists beforehand). Not only could a well-done voice interface be faster, it would be safer.
It's hard to be religious when certain people are never incinerated by bolts of lightning.
It only takes me one on mine[1] - of course, I'm bullshitting - that's if I happen to be in the right folder and have the cursor over the file, but at least I know & admit that I'm bullshitting.
[1] my mp3 player, which is *not* an iPod.
Confucius say, "Find worm in apple - bad. Find half a worm - worse."
Considering both of those are twenty times the size of the controls on my stero, no.
It appears that they were dropped on the head at birth.
Si tacuisses philosophus mansisses. If you had kept quiet, you would have remained a philosopher.
If you're listening to another song, which is likely, then you have to press MENU three times to get back to the artist list, and then the center button three more times (artist/album/song) to select the song. That's six presses total. If you wanted to browse by a different aspect (genre, etc.) it would take eight presses total. It would take four total if you were switching to something from the the same artist, and two if you were in the same album. But if you made a mistake, it would take a couple extra presses to correct. So I think six is typical.
It's hard to be religious when certain people are never incinerated by bolts of lightning.
I'm not sure it's still around, but a player from e.Digital had this feature a few years ago. We had one here for testing.
If you're going through hell, keep going -Winston Churchill
Then maybe you just need to practice more. I can assure you it is possible to operate stereo controls and indeed other electronic gadgets without positioning them in front of your face.
They whose government reduces their essential liberties for temporary security, receive neither liberty nor security.
This company is making a system that can catalog the ID's in music files, and allow you to ask for songs based on genre, title, artist, etc.
While voice recognition has existed for awhile, this is not the same as the system you mentioned. Sure, it's not a big break away in technology, but who said it was.
- It's not the Macs I hate. It's Digg users. -