Electronic device, method of identity verification, and computer-readable storage medium

Abstract

The present invention relates to an electronic device, an identity verification method and a computer-readable storage medium. The electronic device includes a memory and a processor, and an identity verification system is stored in the memory. When the identity verification system is executed by the processor, it is realized: After the current voice data of the target user to be authenticated, the current voice data is framed according to the preset framing parameters to obtain multiple voice frames; the predetermined filter is used to extract the preset type of voice data in each voice frame Acoustic features, generating a plurality of observation feature units corresponding to the current speech data according to the extracted acoustic features; pairing each observation feature unit with the pre-stored observation feature units to obtain multiple pairs of observation feature units; Multiple groups of paired observed feature units are input into a pre-trained authentication model of a preset type to authenticate the target user. The invention can reduce the error rate of short speech recognition.

Description

technical field [0001] The present invention relates to the field of communication technology, in particular to an electronic device, an identity verification method and a computer-readable storage medium. Background technique [0002] Voiceprint recognition is a kind of identity authentication technology through computer simulation and discrimination of the target voice, which can be widely used in the Internet, banking system, public security and justice and other fields. At present, the traditional voiceprint recognition scheme uses a general background model based on Gaussian mixture model to record the speaker, and conducts difference analysis, then extracts voiceprint features according to the difference, and scores them through similarity measure to give the recognition result. This voiceprint recognition scheme has a low recognition error rate for long recordings (for example, recordings with a duration of 30 seconds and above), and a good recognition effect. recor...

AI Technical Summary

Problems solved by technology

This voiceprint recognition scheme has a low recognition error rate for long recordings (for example, recordings with a duration of 30 seconds and above), and has a good recognition effect. recording), due to limited parameters, the subtle differences in the recording cannot be well modeled by using the general background model framework, resulting in poor performance for short speech recognition and high recognition error rate

Images