paper: Addressing Text-dependent Speaker Verification Using Singing Speech
This document is intended to introduce the system and data of the paper. Three speaker recognition systems are respectively under the corresponding files
1.GMM system
GMM model based on Alize platform
2.X-vector
The X-vector neural network model has a Readme file on the Kaldi platform
and the corresponding file, which shows how our Kaldi experiment.
3.DTW
The DTW model is in the corresponding folder.
For the convenience of use, I used some Chinese comments in it.
When used, the corresponding train folder and test folder correspond
to training data and test data respectively.
In addition to the three model introductions, we have also published our database on the Zenodo , the detailed address is: https://zenodo.org/record/3241566 Data description in the data folder.
Then, give the results of all trials files and corresponding model generation.