And now, from the people who gave you the I-mode Internet phone, we bring to you a speech-recognition system so revolutionary that it is, in fact, speechless. NTT DoCoMo showed the prototype system during a rare tour of its Yokosuka, Japan, R&D center.
The system uses electromyography, or EMG, to measure the electrical activity in facial muscles used when a person speaks. Mouthing the words without sounds means the user of such a system could communicate in a noisy room.
For the system to work, three electrodes have to be touching certain areas of the face to measure the electrical activity. In a demo, a user had sensors mounted on his thumb and first two fingers (see photo). The thumb was placed under his chin, the forefinger was held vertically touching his cheekbone, and the second finger held just above his top lip. This strange pose doesn’t impede the mouthing of words.
After three years, the prototype can recognize the five Japanese vowels. Developers are also working on other languages, says Tomoyuki Ohya, director of NTT DoCoMo’s Multimedia Signal Processing Lab. But first, they have to tackle Japanese consonants.