VBHMM x-vectors Diarization (aka VBx)

Diarization recipe for CALLHOME.
The recipe consists of

• computing x-vectors
• doing agglomerative hierarchical clustering on x-vectors as a first step to produce an initialization
• apply variational Bayes HMM over x-vectors to produce the diarization output
• score the diarization output

More details about the full recipe in
F. Landini, J. Profant, M. Diez, L. Burget: Bayesian HMM clustering of x-vector sequences (VBx) in speaker diarization: theory, implementation and analysis on standard tasks

Usage

I run the recipe, execute the run scripts for the different datasets with the corresponding parameters. Please refer to the scripts for more details. The CALLHOME recipes require the corresponding datasets and the paths need to be provided.

This repository has x-vector extractors already trained to function as a standalone recipe. However, the recipes for training the extractors can be found here.

Getting started

We recommend to create anaconda environment

conda create -n VBx python=3.6
conda activate VBx

Clone the repository

git clone https://github.com/Sangramsingkayte/VBx.git

Install the package

pip install -e .

Initialize submodule dscore:

git submodule init
git submodule update

Run the example

./run_example.sh

The output (last few lines) should look like this

File               DER    JER    B3-Precision    B3-Recall    B3-F1    GKT(ref, sys)    GKT(sys, ref)    H(ref|sys)    H(sys|ref)    MI    NMI
---------------  -----  -----  --------------  -----------  -------  ---------------  ---------------  ------------  ------------  ----  -----
demo              7.77  38.23            0.80         0.84     0.82             0.77             0.72          0.60          0.42  1.37   0.73
*** OVERALL ***   7.77  38.23            0.80         0.84     0.82             0.77             0.72          0.60          0.42  1.37   0.73