Unlock instant, AI-driven research and patent intelligence for your innovation.

Vocal print recognition method oriented to smart mobile device

A mobile device, voiceprint recognition technology, applied in speech analysis, instruments, etc., can solve the problem of high computational complexity, achieve good recognition effect, and reduce complexity.

Active Publication Date: 2015-03-25
BEIHANG UNIV
View PDF5 Cites 7 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] Aiming at the problem that the current existing voiceprint recognition technology has too high computational complexity in the recognition process and is not suitable for direct application to mobile devices with relatively limited computing resources such as smartphones, the present invention proposes a voiceprint oriented to smart mobile devices. recognition methods

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Vocal print recognition method oriented to smart mobile device
  • Vocal print recognition method oriented to smart mobile device
  • Vocal print recognition method oriented to smart mobile device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0019] The present invention will be further described in detail below in conjunction with the accompanying drawings.

[0020] A voiceprint recognition method for smart mobile devices (K-UBM for short) of the present invention uses voiceprint recognition on mobile devices with relatively limited computing resources such as smartphones as an application scenario. The time complexity of the voiceprint recognition algorithm is reduced as much as possible under the premise of accuracy, which includes two stages: speaker model training and target speaker recognition.

[0021] First, the speaker model training stage

[0022] Such as Figure 5 As shown, the UBM (Universal Background Model) model used in the method of the present invention is a speaker-independent high-order VQ (Vector Quantization) model trained from a large amount of background speech data including various types of speakers. It mainly includes three processes of preprocessing the background speech, feature extrac...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a vocal print recognition method oriented to a smart mobile device. Vocal prints on mobile devices with relatively limited computing resources such as a smart phone serve as an application scene, the method aims at reducing the time complexity of the vocal print recognition method as much as possible on the premise of not affecting the identification accuracy, and the method particularly comprises a speaker model training stage and a target speaker recognition stage. In the speaker model training stage, significance of codons in each speaker VQ codebook is reflected on distribution of weights; in the vocal print recognition stage, only extracted K codons with the maximum feature vector and weight are matched, and therefore the computing complexity of the system is reduced effectively, and the recognition rate of the system is improved.

Description

technical field [0001] The invention belongs to the field of voiceprint recognition, and in particular relates to a voiceprint recognition method for intelligent mobile devices. Background technique [0002] Voiceprint recognition (also known as speaker recognition) technology is a technology to identify the identity of the speaker to be tested according to the voice parameters in the voice waveform that reflect the speaker's physiological and behavioral characteristics. The research on voiceprint recognition began in the 1930s, and the early work mainly focused on the experiment of hearing and distinguishing speech with the human ear and exploring the possibility of listening recognition. With the development of electronic technology and computer technology, it becomes possible to automatically recognize human speech by machines. Pruzansky of Bell Labs proposed a voiceprint recognition method based on pattern matching and probability analysis of variance, which attracted t...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G10L17/02G10L17/04
Inventor 牛建伟刘闯王磊王博伟
Owner BEIHANG UNIV