A fusion method and device for voiceprint features

Abstract

The disclosure provides a voiceprint feature fusion method, including: extracting voice spectrum features, using the voice spectrum features as input, using a general background model and a global difference space matrix to extract i-vector voiceprint features; using deep neural network Network, extracting x-vector voiceprint features and d-vector voiceprint features; using the i-vector voiceprint features, the x-vector voiceprint features and the d-vector voiceprint features as samples, based on linear discrimination The fusion of the voiceprint features is completed through analysis. By introducing a method based on linear discriminant analysis to fuse multiple voiceprint features, the complementarity of multiple voiceprint features and the discrimination of fusion features are improved, which can ensure that the target speaker can pass through voiceprint authentication. rate, reduce the misrecognition rate of non-target speakers, and improve personalized user experience.

Description

technical field [0001] The present disclosure relates to the field of speech recognition, and in particular, to a method and device for fusing voiceprint features. Background technique [0002] At present, with the popularization of information technology, automatic speech recognition technology is playing an increasingly important role, and its application prospects are also broader. The speech signal mainly contains three aspects of information: who is speaking, what language is spoken, and what is said content. The automatic speech recognition technologies involved are: speaker recognition, language recognition and semantic recognition. Speaker recognition technology, also known as voiceprint recognition, mainly studies the technology of authenticating the identity of the speaker according to the input voice signal. Like other recognition technologies, speaker recognition recognizes the input speaker audio through certain features, so as to confirm the identity of the in...

AI Technical Summary

Problems solved by technology

However, the above-mentioned simple method of calculating the mean value of features or similarity scores to realize the fusion of voiceprint features does not fully consider the complementarity between features and the discrimination of fusion features.

Images