Scientists from the Tomsk University of Control Systems and Radioelectronics have created a mathematical model that recognizes speech and surpasses analogues in certain indicators.
Innovators note that this model could become the basis of a comprehensive personal identification system, as well as facilitate the management of “smart” technology using voice.
According to experts, a method of identifying a person through speech recognition is being developed using digital security technology used to control equipment and security systems in remote banking services.
Scientists from the Tomsk University of Control Systems and Radioelectronics have created a new mathematical model of speech recognition, which is superior in accuracy to analogues and allows you to correctly identify the speaker.
Anton Konev, Deputy Director of the Reliable Interaction Technologies Competence Center of the National Institute of Communications, says: “The mathematical apparatus differs from its counterparts in its great similarity to the human inner ear when recognizing critical speech frequencies.
The potential mechanisms underlying the current speech recognition systems in this device are only ancillary.”
Experts note that the percentage of errors in recognizing the fundamental tone of speech with the new device is four times less than that of the world’s best analogues, and this makes it possible to better recognize ordinary speech and even singing. It also allows you to communicate with smart devices not through commands, but directly, as in the case of an interlocutor.
Experts assure that the new system will allow to verify the identity of the interlocutor during a communication session, which will completely eliminate some types of fraud.
Source: RIA Novosti