Voiceprint recognition application of ordered sequence similarity comparison method

Abstract

The invention relates to the voiceprint recognition application of an ordered sequence similarity comparison method. A voiceprint recognition process comprises voice characteristic space building, subspace division, training statement characteristic extraction, test statement characteristic extraction and voiceprint sequence comparison recognition. Particularly, the voiceprint sequence comparison recognition is realized by the ordered sequence similarity comparison method, and comprises the following steps of: 1) converting sequences into characteristic arrays; 2) comparing the column numbers of the two arrays; 3) performing label matching on the two arrays; 4) calculating the contributions of each label to similarity; and 5) calculating overall similarity. When the method is used for calculating the similarity of the two ordered sequences, the arrangement sequence information of the sequences is fully utilized, so that the method has the advantages of saving storage resources, along with a small amount of calculation and the like, and is applied to an embedded system with finite system resources, and the problems of a probability-statistics-based modeling method are solved.

Description

technical field [0001] The present invention relates to a voiceprint recognition technology, in particular to an application technology that introduces a highly computationally simplified sequence comparison method into a new voiceprint recognition technology. Background technique [0002] Speaker recognition, like fingerprints, iris, face recognition, etc., is a type of biometrics and is considered to be the most natural way of biometric identification, also known as "voiceprint" recognition. Speaker recognition has the advantages of simple acquisition equipment, low system price, and easy acceptance by people. In access control systems, safes, personal equipment (cars, computers, mobile phones, PDAs, etc.) Voice can realize the learning and testing process, which has outstanding application advantages. [0003] The basic process of speaker recognition is speech acquisition, feature extraction, and classification model. The common speech feature extraction method is to u...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to propose a voiceprint recognition application of an ordered sequence similarity comparison method to solve problems such as high computational complexity and dependence on sequence length

Images