Implementation of the Standard I-vector System for the Kaldi Speech Recognition Toolkit
Year: 2016
Type: article
Abstract: This report describes implementation of the standard i-vector-PLDA framework for the Kaldi speech recognition toolkit.The current existing speaker recognition system implementation is based on the Subspace Gaussian Mixture Model (SGMM) technique although it shares many similarities with the standard implementation.In our implementation, we modified the code so that it mimics the standard algorithms in the ivector based speaker recognition system.The implementation is compared with the existing Kaldi recipe for speaker recognition on the NIST SRE 2008 evaluation set.The entire implementation is made available at https://github.com/idiap/kaldi-ivectorunder the Apache 2.0 license.
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W2587848299
