Researchers at the Massachusetts Institute of technology (MIT) invented a system called AlterEgo, is able to recognize non-verbal cues — that is, in fact, read minds media. The development consists of computer systems and devices, which is attached behind the ear, passes through the line of the chin and ends near the mouth of the user.

The electrodes of a wearable device catch neuromuscular signals in the jaws and face, which creates inner speech — the words that the person says in

my head. These signals are passed to a machine learning system that analyzes data and connects some signals.

The system can also interact with the user through the “kastorovogo” earphone, transmitting of vibration from the face to the ear. It is intended that the carrier was comfortable to obtain the desired information without interrupting the conversation.

The researchers tested the system in various situations, including playing chess, multiplication and addition, using in each case a lexicon of 20 words. With these 20 words, the design achieves a precision of 92 %.

Over time, the researchers want to reach the accuracy of 100 %. Another example of using AlterEgo — select a movie to watch through control over what is displayed on the TV screen.

To create such a device, it was necessary to determine the location of the facial points with the most reliable neuromuscular signals. For this experimental set 16 of electrodes in different parts of the face and asked four times to pronounce in mind the sequence of the same words. So it was found seven of the most effective areas. Now the employees of the Institute are trying to create the same device but using only four points on the jaw line.


Leave comment

Your email address will not be published. Required fields are marked with *.