h a l f b a k e r yThe best idea since raw toast.
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
[UPDATES at the end]
Most new laptops,netbooks come with integrated webcams on the top. I can immediately think of two really usefull ways of using this.
1) Replace input device activity detector with face detector for sensing user presence.
[This idea is now BAKED in Samsung Galaxy S3,
and the feature is called SmartStay ]
Instead of relying on keyboard and mouse activity, rely on availability of face for activity sensing. No face in front for more than 1 minute? lock screen, show as away on IM. Face came back ? show "unlock" password prompt. This is especially usefull if you are watching a movie or reading a long article using auto-scroll. The screen won't lock/go to screen saver mode in the middle of it.
2) Auto tilt.
I sometimes use my laptop when I sleep on my side to read really long pages. As I use GNU/Linux, I just issue the command xrandr -o 1 or xrandr -o3 depending on which side I am sleeping on. As far as I can remember many recent Windows graphics drivers, even some old ones register Ctrl+Alt+cursorkey to change orientations.
My idea was to use the web cam to detect which angle your face is on.
There are enough face detection softwares out there. Even good open source ones. If someone who can understand all the code can make it extract the angle of the face, then the angle can be fed to another piece of code which uses simple heuristics and time outs to change the orientation of the screen when the user is seeing the screen at 90 degrees.
3) Auto-brightness
At least in linux, its quite easy to change the brightness of the laptop screen, you just write your brightness level to a special file and done. if you can find out the ambient brightness from the camera, then just simply adjust screen backlight brighness appropriately through code. done.
4) Could also be an interesting input for gaming. Though it can be a pain in the neck if the game gets too continuous or fast paced. But the possibilities are interesting.. [a] In a FPS you can use it to control the head of the charector (to look around). [b] In a flight simulator, you could use it to control the pitch, roll, etc
Even better: If you look away from the screen for more than 2 seconds the game can pause itself!
5) [ok.. ideas are getting a bit crazier now]
Hands free! Use it for alternative input for dialogs.. Nod your head around x axis for yes. nod your head around y axis for no. Hands free!
Hands free browsing!. Firefox plugin and extension which can be switched on while having lunch. The plugin that knows where you are looking and automatically scrolls it to centre (very very useful if you have the habit of reading slashdot or other news while having food). A left wink can bookmark the page and position for rereading. A right wink can bookmark the page and position for commenting later on(after lunch).
6) Sleep pause
If you are watching a movie and doze off, the movie should get paused, and the laptop can go to low power mode. ( I am confused whether I should be really adding this idea in the "contributed" section as i got the idea from [bigsleep]'s nick :-)
7) Boss / kids / wife detector.
Saves a lot of embarrassing moments ;-)
8) Guesture based reading of cards (Visiting cards, barcodes(?) semacodes, etc)
If I point at an object (near the camera) edge with an index finger, it should try to find text/barcode/semacode on it. OCR it and present in a dialog. After I confirm or correct and then confirm the input, operations follow suit
- URL is bookmarked/launched, visiting card info is stored in my addressbook, etc.
The cam of my EEE Pc is 1.3 megapixel and so I guess all newer notebooks carry good enough cameras.
Contributed ideas:
[2 fries shy of a happy meal]
".. detect how far a persons face is from the screen and change font size accordingly.."
[bigsleep]
".. if one of a number of faces disappears[while watching a movie] it could press pause."
[supershnitzel]
augment speech recognition with face detection. just like hearing-impaired people do.
[UPDATE 8: Idea (2), (4)+(6) have been baked in Samsung Galaxy S4
In addition (5) seems to be somewhat baked (at least solves the same problem) . And 21Quest's suggestion of Air Gestures have also been implemented.]
[UPDATE 7: Idea (1) is now baked!! in Samsung galaxy S3]
[UPDATE 6: Since I posted this, I have regulary come across stuff that uses head tracking, every time increasing my hopes that all these ideas are coming closer to reality .. See links]
[UPDATE 5: added another idea 8) ]
[UPDATE 4: modified idea 7]
[UPDATE 3: added another idea 7) ]
[UPDATE 2: added one more contributed idea and changed main title and title of 1) of so that Jutta does not get thrown off ;-)]
[UPDATE: added 2 more ideas 4 and 5. also added a contributed idea]
Dual spectrum Eye Recognition
dual-spectrum_20eye_20recognition
got any laptops with IR cams ? [kamathln, Feb 26 2009]
(5) implemented by opera!!!
http://my.opera.com...log/face-the-future
(seems to be april fool though :-( ) [kamathln, Apr 01 2009]
CamSpace
http://www.youtube....watch?v=v0srY37kkMw
Comes close to (4) [kamathln, May 02 2009]
Sixth Sense
http://www.ted.com/...he_sixth_sense.html
Comes close to 8 [kamathln, May 02 2009]
smert.net
http://smert.net/20...-beta/#comment-2805
Implementation by Jason [kamathln, May 02 2009]
Presence sensor *almost* baked!
http://www.pcworld...._cant_get_here.html
not on sale in the US yet though... [DIYMatt, May 20 2009, last modified May 21 2009]
Mousetrap
http://live.gnome.org/MouseTrap
Gaze based cursor control [kamathln, Jul 02 2010]
John Underkoffer's hand gestures demo
http://www.ted.com/...with_a_gesture.html
21 Quest 's suggestion :baked? Just have to adapt it to Laptop :) [kamathln, Jul 04 2010, last modified Jul 06 2010]
same UI
Chat_20Enhancer
for text-based conversation [FlyingToaster, Jul 04 2010]
HandVu
http://www.movesins.../HandVu/HandVu.html
Hand Gesture Recognition using OpenCV [kamathln, Jul 07 2010]
Auto-dimming TV
http://www.gizmag.c...watching-you/15330/
[comes close to (1), but on tv and not on a computer. Sony Bravia TV dims/switches off when you are not looking [kamathln, Sep 15 2010, last modified Sep 13 2012]
Ubuntu getting physical
http://design.canon...9/getting-physical/
[kamathln, Sep 16 2010]
Compiz head tracking
http://ubuntuforums...thread.php?t=952895
[kamathln, Sep 16 2010]
Proxemic media player concept
http://www.youtube....watch?v=OHm9teVoNE8
Comes close to the string of ideas in comments that came about thanks to idea #6 [kamathln, Oct 01 2010]
SmartStay of Samsung Galaxy S3
http://www.samsung....xys3/smartstay.html
1) is baked and the most talked about feature in Samsung galaxy S3 [kamathln, Sep 04 2012]
Please log in.
If you're not logged in,
you can see what this page
looks like, but you will
not be able to add anything.
Destination URL.
E.g., https://www.coffee.com/
Description (displayed with the short name and URL.)
|
| |
My powerbook does the brightness level thing. |
|
| |
[+] suggest that when the webcam is facing the operator it operates as a webcam, but operates as a brightness adjuster while facing outwards... |
|
| |
bigsleep: yeah.. big sleep is exactly what I need. thanks for pointing out. |
|
| |
Very good.
Could it also detect how far a persons face is from the screen and change font size accordingly? |
|
| |
[2fries] Oh Yeah. But I guess that option should be disabled by default. But yeah, fantastic idea. It could surely do that. You could also write a compiz plugin to automatically "zoom current window" as an alternative. |
|
| |
With the zooming, you could go two ways - you could interpret the user moving closer as "I want to have a closer look, make the font larger" (exaggerating what would subjectively happen normally); or you could interpret it as "I don't need to make the font large anymore, because the user's head is closer" and *shrink* the font size. |
|
| |
(1) - not sure if that works. [my misunderstanding of presence sensor as motion sensor --ju]
(2) - very clever. Wish I'd thought of that!
(3) - might be a good first stab, but in the long run, I really want the brightness to adjust to what's behind the screen, not to what's in front of it. E.g., if I'm sitting before a window, and there's sunshine out, I need the screen to be brighter. |
|
| |
jutta: You are right on it with 3) .. But I guess you missed the point with 1) (might also be my improper communication). The idea with 1) is, if your face is available anywhere in the image caught by the camera, you are using the computer. You need not move your face. Just the presence of a face means someone is using the computer. You should be able to set different distances for different purposes. If you are watching a movie, a face detected at any distance can make the computer feel someone is using it. There can be a default minimum distance for everything else which makes the computer feelsomeone is using it(otherwise someone, even unintentinally staring at the laptop in the office could trigger the "unlock screen" login screen.) |
|
| |
// Just the presence of a face means someone is using the computer.
Duh! That makes sense, of course. I guess the "activity sensor" title threw me off - the explanation itself is completely clear, on rereading. |
|
| |
[bigsleep] your nick gave me one more idea. If there is only one face and it falls asleep, it should press pause as well. And optionally trigger a negative vote to that movie(okay, just kidding). |
|
| |
Did i make history on .5B by getting an idea from someone's nick ? ;-) |
|
| |
I suggest- augment speech recognition with face detection. just like hearing-impaired people do.. (...scuffles to uspto) |
|
| |
supershnitzel : I actually did not get what you said, but I got an idea from it, which is a totally different idea altogether. If the webcam can do a bit highres at good FPS, then it must be able to look at your lips and deduce your lip muscle movements. This can be used by speech recognition in addition to normal microphone, to get more accurate results. |
|
| |
kamathin - that's exactly what I meant. What hearing impaired people do. it's called "reading lips", augmenting (=improving) hearing with vision. or in computer terms, augment speech recognition with face detection. |
|
| |
and I felt I had gotten a totally different idea ! Wierd! |
|
| |
[up_on_cloud_nine] The way most people do when they agree. With "on" X axis, I meant take the X axis as the axis around which you turn your face. I was thinking in terms of how the computer would know about it. So mentioned in geometric terms :-p |
|
| |
[+] With gaze detection, left-wink and right-wink generalised to the whole UI (and not just the browser), you have full mouse-equivalent control of your computer. |
|
| |
Though I don't know the consequences of sudden sneezes. |
|
| |
Computer; mouse cursor on "Edit" |
|
| |
computer: "delete: Are you sure to delete all files in /home/yourname" |
|
| |
<uncontrollable third sneeze/> |
|
| |
//How do you nod your head on the x axis???// |
|
| |
For the international community that is a quite correct way to descibe the motion - in Bulgaria atleast when you nod (y axis) it means no, and when you shake (nod on x axis) it menas yes.... VERY confusing for a western european if you've not been informed before hand:
visitor "is this the way?"
local "shake head" (smile)
visitor "so which way is it?" (smile an confused)
local "that way! more shaky head" (energetic smile and shake)
visitor "not that way? nodding head" (more confused)
local "No it is that way, nod head at your shaky head! and shake head at pointing direction" (tiersome but still energetic smile)
visitor "IS IT THAT WAY OR NOT?" shake and nod for appriopriate word being shouted
local "YES, shake head!"
etc.... |
|
| |
I imagine there must be other countries that this also applies to.. |
|
| |
aah text get bunched up, sorry for not so easy to read post...
Love the idea, webcams surely could come with little add-ons like these with out much effort! like the pause while sleeping during movie... save having to watch bits you thought you hadn't seen untill 10mins later when you get to bits you remember!! |
|
| |
strayllamma. I have first hand experience .. I am from India. A German guy had been to our company for his internship. After about a month of confusion, he openly expressed it. Its after a day of brainstorming that I got an idea of how to explain it to him. |
|
| |
Actually he was not that confused by yes and no. We Indians have one more : "Ok" or "comply". To express that you nod your head around the Z axis. You read that right, as if the Z axis goes through your nose or neck (changes fuzzily depending on the situation). |
|
| |
guy1 : Hey.. close the door before you leave, Okay? |
|
| |
guy2 : <nod what="head" axiz="z">Okay</nod> |
|
| |
Mom: Whenever you sneeze, you should keep a hand kerchief in front of your mouth to prevent a nasty situation. |
|
| |
Little Kid: <nod what="head" axis="z">Okay mom</nod> |
|
| |
Mom: Have you seen my spectacles anywhere? |
|
| |
Son: <nod what="head" axis="y"> No mom</nod> |
|
| |
Whew! I think this should go into Wikipedia or something. LOL |
|
| |
This idea wouldnt be very difficult to create right now. I know i can implement most of them into a workable program in 3 weeks time. It's an interesting tool. If there was any hope of it making back small donations from users (i think most people will just want to use it as free) then i'd make it for you. |
|
| |
quantass: actually, I myself have been thinking of learning and implementing it as a LGPLed library and app, implemented in PyCV. But unfortunately, I got enough to work on already and cant afford to invest time. So I thought It would be best if I let out the ideas on halfbakery for free, so that if there is someone who can implement it as a LGPL/GPLed app, the idea should not be dormant in my head. |
|
| |
You can probably do this as a dual licensed app. GNU/Linux users would get it for free(as in both price and freedom). If you can market it in a nice way, other OS companies (both self-made and integrators) might want to buy exclusive code licenses (thanks to dual licensing). You might want to check out how MySQL makes money from their dual licensing schemes. Also, if your app was real good, and setup a paypal account, you may even get a load of donations! |
|
| |
Disclaimer: I am not a lawyer. Soo you may want to consult a good FOSS lawyer before doing anything. |
|
| |
Is this the longest-titled idea on this website? |
|
| |
LOL! Maybe. This website surely has one of the longest (though surprisingly manageable) query form. Even then, you cannot search/sort by title length. Otherwise we probably would have found out. |
|
| |
[bobofthefuture] Glad to be usefull, but please explain what was the problem and whats the solution you got. |
|
| |
Idea title length: this one is one of the ten longest, but not the longest - that achievement goes to "Shopper's baseball cap with 2 way radio, quadrature direction finding antennae and head-up projection display." |
|
| |
Whoa! thanks for the info Jutta! |
|
| |
This just got *almost baked*! At least, the face-based presence sensor part. Available in Japan. See link. |
|
| |
Whoot! They just took an infrared intruder detector security alarm and stuck it into a monitor! Wise thinking! |
|
| |
I also wish it could be connected to the CPU via USB. Along with the infrared pixels, it could also push different info about where objects were detected, and what the code on the monitor thinks about it. That way, softwares could use it. |
|
| |
[DIYMatt], It comes close, but it is still not baked. |
|
| |
1) The camera in the product you link to controls only the monitor. |
|
| |
For my idea to be baked, the information must be transferred to the computer and the operating system/ applications must be able to use the information for sensing Human face presence and take necessary actions. |
|
| |
2) The "camera" only detects Infrared in the range of human beings or mammals and not a human face. From what I collected, It either has a single infrared diode or a simple 100px by 100px sensor. |
|
| |
For my idea to be baked, a proper face must be recognized by the code. |
|
| |
So I request you to edit the "Baked" text to "Almost baked" |
|
| |
Would the face recognition bit know the difference between my face, the face of Groucho Marx printed on my T-shirt, and the face of someone peering over my shoulder? |
|
| |
@Twizz: At least in the current state of the idea, there is no face recognition(whose face is it?), only face detection(is there a face). So the answer is "no, it cannot". |
|
| |
With 2 cameras (thus percieving third dimension), it could probably avoid getting confused by "the face of Groucho Marx printed on my T-shirt". It could always notify you subtly when the number of faces increases/decreases. |
|
| |
either that or just look for the most centered face to take cues from... teach it to read lips. |
|
| |
FlyingToaster: Good one.. equivalanet to "secret knock" |
|
| |
@21 like that ? [check link] |
|
| |
@21 Quest: I wish someone could revive HandVu [link] |
|
| |
If you are wondering why this idea comes back as updated every now and then, check the links section. |
|
| |
Hear ye Hear Ye! Idea (1) got baked! check links |
|
| |
Samsung > Galaxy > S3 > SmartStay |
|
| |
Hear ye Hear Ye! Idea (2), (4)+(6) and (5) also got baked! |
|
| |
Samsung > Galaxy > S4 > SmartRotate, Smartpause and SmartScroll |
|
| |
Smart scroll is not exactly what we described here though. |
|
| |
In addition, "Air gestures" suggested by "21Quest" on "Jul 04 2010" has also been implemented. |
|
| |