A Georgia Tech team has built a working app for latest-generation mobile phones that uses the built-in accelerometer to guess which words you're typing on your PC's keyboard, by measuring the movements of your body as you type.
The technique works through probability and by detecting pairs of keystrokes, rather than individual keys (which still is too difficult to accomplish reliably, Traynor said). It models "keyboard events" in pairs, then determines whether the pair of keys pressed is on the left versus right side of the keyboard, and whether they are close together or far apart. After the system has determined these characteristics for each pair of keys depressed, it compares the results against a preloaded dictionary, each word of which has been broken down along similar measurements (i.e., are the letters left/right, near/far on a standard QWERTY keyboard). Finally, the technique only works reliably on words of three or more letters.
For example, take the word "canoe," which when typed breaks down into four keystroke pairs: "C-A, A-N, N-O and O-E." Those pairs then translate into the detection system's code as follows: Left-Left-Near, Left-Right-Far, Right-Right-Far and Right-Left-Far, or LLN-LRF-RRF-RLF. This code is then compared to the preloaded dictionary and yields "canoe" as the statistically probable typed word. Working with dictionaries comprising about 58,000 words, the system reached word-recovery rates as high as 80 percent.
(via /.)
(Image: Keyboard, a Creative Commons Attribution Share-Alike (2.0) image from bull3t's photostream)
