Startup SpeakIn from Chinese Shenzhen is working on a system that determines the unique characteristics of a person's voice and compares them with a database of voices. As the South China Morning Post writes, technology needs a few seconds to identify a person.

"Voice prints" are completely personalized and can serve as an additional biological identifier, hightech.fm reports. The technology will be useful first of all for the creation of security systems. For example, she will be able to recognize and track criminals: a bank robber can hide a face behind a mask, but distorting the voice will be much more difficult. Currently, SpeakIn is working closely with the Celestial Police to create a database of votes.

It is claimed that the accuracy of the technology created by the company is 98%. In addition to being used by the police and the government, it can be used in smart home applications, to run cars or authorize users, for example, in banks. Today, many devices respond to voice commands but are not able to find out who is talking to them. Using "voice fingerprints", you can teach them to do it. As a result, for example, the TV will include a specific channel for each user. And if your child wants to start the car with a voice command without your knowledge, the system will recognize it by voice and block the operation.

SpeakIn was established in 2015 and has already attracted a number of large investors. Currently, Lenovo, Tencent, Asus, China Merchants Bank and an unnamed automaker are cooperating with the company.