A huge number of gadgets support voice control. This is quite convenient, but the technology has a problem: loss of privacy, problems with recognizing a spoken command in a noisy place, devices often make mistakes and do not show up in voice owners. Group of enthusiastic people

The researchers started from one important observation. The fact is that with the silent passage of the shovel, the movement of bones and muscles provokes a deformity of the owl prodod. With this model of deformities, it can be compared with the formulas of the phrase. The EarCommand device, located close to the headphones, is a miniature speaker. It generates and directs waves close to ultrasound into the ear canal. A sensitive microphone senses an “echo” reflecting off the channel walls. The signal is then analyzed by a computer.

The technology has already been tested. Scientists have developed 32 commands in the form of rare words and another 25 in the form of phrases. The quality is stably stable: the number of errors did not exceed 10.2 and 12.3%, son.c These indicators can be improved as the observation approaches. The experimental system works well at high noise levels and when playing as a face. The developers consider it a PERSPECTIVE and continue study.

Source: Tech Cult

Previous articleFlights without passengers causing unnecessary pollution in the UK
Next articleGoogle Pixel Watch vs Apple Watch Series 8: Which is the best option?

LEAVE A REPLY

Please enter your comment!
Please enter your name here