Slashdot Mirror
Mouse May be Replaced by "Nouse"
mteichrob writes "A Canadian inventor has designed a computer mouse steered by movements of the nose and eyelids. The invention, dubbed a "Nouse," is meant to help people with a disability use a computer.
I think this could also be quite useful for the average office lackey like me! As good as my ergonomic mouse/keyboard combo is, less body movement is still a better replacement."
A feeling of having made the same mistake before: Deja Foobar
If it's the same one (I couldn't access the CBC report linked to so I googled it) then it's much more than just using your nose and eyelids for a mouse- it's full face tracking software at a subpixel resolution.
Which is a really cool idea- it means that a webcam can replace the mouse.
SJW: a person who perceives an injustice, and while correcting it, commits a greater injustice.
http://www.atarimuseum.com/videogames/consoles/260 0/mindlink.html
--- Ban humanity.
You can find Dmitry (the inventor's) webpage here. Lots of interesting stuff, including some interesting facial recognition and tracking work.
Here
I don't have a web-cam handy, so if someone could give this a spin a let us know how well it works, that would be very much appreciated.
If it works well enough, I would definitely use something like this to replace my mouse for everyday tasks. It wouldn't replace the keyboard, but it's always been the mouse that's caused most of my RSI-like symptoms.
Please mod this post only if you think others should/n't read this. I have enough ego^H^H^Hkarma. Thanks!
Prof. Tom Hutchinson at the University of Virginia has been researching this subject for decades. His system lets you control the cursor with the movement of your eyes and can track & record your eyes as you look across the screen. It's been in use for years for a great deal of UI research in addition to it's potential for handicapped users.
I'm no longer current on the research on pointing and control devices, and I don't recall reading any of Gorodnichy's papers, so for what it's worth:
This approach, at it's best, could work as a secondary pointing and control device. I primary device requires a very high degree of control. The hands and fingers are excellent for manipulating a primary device because they are very good at precise, controlled movement and manipulation of other objects through tactile feedback alone and especially well in coordination with vision.
The nose and eyelids, in contrast, are not capable of such movement and coordination. People don't have the necessary muscular control and coordination in their neck and facial muscles.
The human body does have some alternatives to the hands that can be good alternatives to those who cannot use their hands and fingers: feet, toes, tongue, and breath are the most promising.
This is nothing new. Alternatives to the physical hardware with which we interact with our computers has been researched for years. I haven't rtfa, but I'm going to assume that the nose is tracked and associated with the mouse, and two blinks == one click, and three blinks == double click. I did this over two years ago, and I know alot of others who have too, just go to any university. Anyway, tracking the nose is good, but I found tracking your irises(is that even a word) easier and significantly more accurate. I found it easier because you are already focusing on that area for blinks, you already know where the eyes are so there is no need to go searching for other features. Also, the eyes stick out significantly more than a nose, and depending on the lighting, a nose can be very hard to detect because it is skin and blends with your face. I found using the eyes for mouse movement more accurate because you have redundancy, instead of hoping that you found the nose (which you only have one of) and then hoping that you tracked it right, using the eyes you can double check all your answers to a certain precision (doesn't work with a "dead" eye). Also, the iris is clearly defined and surrounded by white in all people. Using the nose, or mouth, for tracking was awkward because as you move your nose up or down you have to move your eyes opposite of that direction to stay focused on the mouse. When you use your eyes, they naturally are pointing where ever the mouse is, and thats most likely your intended area of focus. One final note, as this is getting long, some people may be wondering why 2 blinks = 1 click, etc... Well at first I did have 1 blink == 1 click and 2 blinks == 2 clicks. Debugging this was hell! It drove me literally insane. My mouse kept clicking when it wasn't supposed to and I had no idea why. Then after quite a while of damn near mathematically proving my code to be correct, I realized it was me inadvertenly blinking! Now your thinking, "Duh!", but when its late at night and your on a roll coding and testing, you dont really pay attention to when you blink, I mean we blink all day long and never realize it. So inorder to effectivly click with the eyes, you need to add an extra blink. Just my 2 cents. Oh yea, I also am unfortunately no longer working on that project(I was going to release it on sf.net or something) but around the same time I migrated entirely to linux and my damn Creative Camera didn't have drivers. I'm still waiting to get a logitech cam, any recommendations?
Regards,
Steve
p.s. I forgot to mention that tracking the nose is useful for one application that me and friends found, if you make a triangle with your nose and eyes then you can get a pretty good idea of the orientation of someone's head in relation to 3 dimensional space. Using the eyes isn't very useful for more then 2 dimensions.