In the future, we don’t need to speak. MIT students in the United States recently published a current signal that can be read through the muscles of the facial nerve system, which translates into a speech reader. The researchers said that the average recognition success rate reached 92. %, Prior to this, the researcher tested the use of this “Reader” and successfully ordered pizza without saying a word and his finger was not moving.
The device is called “AlterEgo” and was produced by Arnav Kapur, a graduate of the Media Lab at the Massachusetts Institute of Technology. His work uses a 3D printed plastic device like a white banana. On one side of his face, in the absence of a voice or hand, he can turn a TV, change the color of smart light bulbs, play chess, solve complex mathematical problems, and even call Pizza.
AlterEgo itself can not really "read", but it is close to the user's ear until the lower jaw, using its four sets of electrode sensors to obtain the weak facial muscle current signal of the nervous system. The user is ready to say a When the English word is used, the AlterEgo system can sense the current and convert the current into the corresponding text through the artificial intelligence translation system and send it to the computer.
The researchers found that the average accuracy of AlterEgo's test can reach 92%. At present, it can identify 0 to 9 digits and about 100 English words. After adding the font in the future, it can correspond to more speech content.
AlterEgo also paired with bone conduction speakers to convert external sounds into skin and ear bones and transmit them to the inner ear hearing system. The developer Capria said that he hopes to use this “reader” in the future so that people in need can fully communicate with the outside world.
