Keystroke Biometrics and Different Keyboards

[Juiceion]

Hello guys, I am currently doing some research on Keystroke Biometrics. If you guys don't know what that is it is basically a backend keyrecorder that takes your typing signature (time keys are held down, which shift you use for which letters, time in between keypresses etc...) and uses it in order to authenticate your login. So basically if your signature is not present when you type your password, then the system will lock you out regardless of if you got the password correct.

I am mainly focusing on the different types of key switches and how it effects the verification process, and need some information in order to figure out the remainder of my research.

I really need 2 things: The activation point of Topre on an HHKB preferably in graph form, and if you guys had any suggestions for this that I can add to my presentation.

So far I have a board with custom switches (Holy GSus), a Topre board, a really old MX Blue Board, a scissor switch board, and access to the basic corporate HP/Dell rubber dome desktop board. I have decided to focus on a couple different aspects, including keylogging, remote accessing (using a keyboard that is connected to the computer via a program like TeamViewer), and mobile access (typing the password in on your phone), but want to bring some more into the paper because I feel like it will not end up being long enough.

Let me know what you guys think!

[tp4tissue]

Most keyboards run 125hz.

How precise a signature can you poll with your method ?

Are these signatures also so varied between individuals with same levels of skill ?

___ I don't doubt you, but I have trouble imagining people who touchtype would have significantly different hold times.

Especially on mechanical, because the mid-actuation point 2mm of most switches, people who adapt to mechanicals, do not hold much at all.

[Juiceion]

Most keyboards run 125hz.

How precise a signature can you poll with your method ?

Are these signatures also so varied between individuals with same levels of skill ?

___ I don't doubt you, but I have trouble imagining people who touchtype would have significantly different hold times.

Especially on mechanical, because the mid-actuation point 2mm of most switches, people who adapt to mechanicals, do not hold much at all.

It is using an already created keystroke biometrics API. If you want to test it with 2 people of similar touch-typing capabilities let me know and I will figure out how to do that. It may include cookie duplication and transfer in order to successfully do if I am unable to set up the API on a placeholder website, but as far as I can tell, the differences even between people of similar typing skill is significantly different and will throw a 0% match.

I am working with a short amount of time and since this is just research for a class, and not something I am getting paid for, I am not overly concerned with the creation of a standalone site, or my own keystroke biometric software. If I was creating my own keystroke biometric software I would have to triple or quadruple my working timeframe. I am still looking for a standalone open source software on github but so far the closest thing I have been able to find is a JavaScript API that is not open source.

But as for the touch typing, it also uses time between keypress, which is going to be different for everyone. And even then on touchtyping it is likely that people still hold at base for different amounts of time. It takes the entire typing signature into account, not just the time that a key is held at actuation.