4 ms·
It seems that the way blind type works is to keep track of the spatial difference between key presses and then compare the typed words to dictionaries to determ
by hristov 16y ago
It seems that the way blind type works is to keep track of the spatial difference between key presses and then compare the typed words to dictionaries to determine a valid word.
There are a couple of problems with this:
First it assumes that you are already used to the keyboard. Thus, even if the letters you are typing are not correct, the spatial offsets between the letters you are typing are roughly similar to the offsets of the correct letters. So basically, it will work when you know the keyboard and know roughly where each letter is in comparison to other letters, you just keep missing the letters. I don't think it will help most people because most people are still not very familiar with mobile phone keyboards.
Secondly, I am pretty sure this is based on dictionary matching (it has to be). That is always a problem because a lot of people that use mobile devices just cannot spell, do not spell or insist on using slang or inventing their own words and their own spellings.
There have been many dictionary based schemes to help people text message and most of them have failed.
- StavrosK 16y agoThe dictionary model is most likely used for calibration, e.g. to get 3 points necessary for the correct orientation of the keyboard. This will likely be adjusted for drift as you type, but if a user presses a r instead of a n I very much doubt the entire KB will be shifted. It will probably just work like a normal keyboard, with spelling correction/suggestion, but one that gets shifted as you type.