Method for starting up mobile terminal through voice password

Abstract

The invention discloses a method for starting up a mobile terminal through a voice password, wherein a user voice print model, a pronunciation content model and a pronunciation rhythm model which are pre-established are stored in the mobile communication terminal. The method comprises the following steps: extracting the acoustic feature from the sound wave of the mobile terminal user and forming a voice print feature vector sequence, a pronunciation content feature vector sequence and a pronunciation rhythm feature vector sequence which are to be identified; matching and comparing the pronunciation content feature vector sequence to be identified with the pronunciation content model one by one, and obtaining a pronunciation content matching score A; matching and comparing the voice print feature vector sequence with the voice print model one by one, and obtaining a voice print matching score B; matching and comparing the pronunciation rhythm feature vector sequence with the pronunciation rhythm model one by one, and obtaining a pronunciation rhythm matching score C; performing operation to the comprehensive matching degree of the pronunciation content matching score A, the voice print matching score B and the pronunciation rhythm matching score C, and starting up the mobile terminal when the obtained comprehensive matching degree Z is equal to or larger than a preset threshold value.

Description

technical field [0001] The invention relates to a method for starting a mobile terminal, in particular to a method for starting a mobile terminal with a voice password. Background technique [0002] In recent years, with the rapid development of electronics and computer technology, there are more and more types of mobile terminals, more and more powerful functions, and lower and lower product prices, including mobile phones, PDAs, MP3 / MP4 and other mobile terminal products. One of people's daily necessities. The portability of mobile terminal products also makes it one of the property that people are most likely to steal and steal. [0003] At the same time, with the development of speech recognition technology, there are more and more commercial applications related to speech recognition and speech synthesis. It is introduced into the mobile terminal as a power-on authentication method to ensure the security of the mobile terminal. [0004] The production of human langua...

AI Technical Summary

Problems solved by technology

[0013] (3) Neural network method: There are many forms, such as multi-layer perception, radial basis function (RBF), etc., which can be explicitly trained to distinguish speakers from their background speakers. The amount of training is large, and the model is reliable. poor promotion;

[0016] (6) Multinomial classifier method: it has higher precision, but the model storage and calculation amount are relatively large;

At the current level of technology, the two cannot reach the minimum at the same time, and the threshold needs to be adjusted to meet the needs of different applications. For example, when "ease of use" is required, the false rejection rate can be lowered. At this time, the false acceptance rate will increase, thereby reducing security; in the case of high "security" requirements, the false acceptance rate can be lowered, and the false rejection rate will increase at this time, thereby reducing the ease of use

Images