STC Speaker Recognition System for the NIST i-Vector Challenge
نویسندگان
چکیده
This paper presents a Speech Technology Center (STC) system submitted to the NIST i-vector Challenge. The system includes different subsystems based on PLDA, LDA-SVM, RBM-PLDA and DBN-PLDA. We propose an original iterative scheme for clustering the NIST i-vector Challenge devset. We also introduce the RBM-PLDA subsystem in the NIST i-vector Challenge. Experiments performed on the progress dataset demonstrate that although the RBM-PLDA and DBN-PLDA subsystems are inferior to the other subsystems in terms of absolute minDCF, in the fusion they provide a substantial gain into the efficiency of the resulting STC system, reaching 0.239 at the minDCF point.
منابع مشابه
The NIST 2014 Speaker Recognition i-Vector Machine Learning Challenge
During late-2013 through mid-2014 NIST coordinated a special machine learning challenge based on the i-vector paradigm widely used by state-of-the-art speaker recognition systems. The i-vector challenge was run entirely online and used as source data fixed-length feature vectors projected into a low-dimensional space (i-vectors) rather than audio recordings. These changes made the challenge mor...
متن کاملNIST language recognition evaluation - plans for 2015
We discuss two NIST coordinated evaluations of automatic language recognition technology planned for calendar year 2015 along with possible additional plans for the future. The first is the Language Recognition i-Vector Machine Learning Challenge, largely modeled on the 2013-2014 Speaker Recognition i-Vector Machine Learning Challenge. This online challenge, emphasizing the language identificat...
متن کاملRBM-PLDA subsystem for the NIST i-vector challenge
This paper presents the Speech Technology Center (STC) system submitted to NIST i-vector challenge. The system includes different subsystems based on TV-PLDA, TV-SVM, and RBM-PLDA. In this paper we focus on examining the third RBM-PLDA subsystem. Within this subsystem, we present our RBM extractor of the pseudo i-vector. Experiments performed on the test dataset of NIST-2014 demonstrate that al...
متن کاملSummary and initial results of the 2013-2014 speaker recognition i-vector machine learning challenge
During late-2013 through early-2014 NIST coordinated a special i-vector challenge based on data used in previous NIST Speaker Recognition Evaluations (SREs). Unlike evaluations in the SRE series, the i-vector challenge was run entirely online and used fixed-length feature vectors projected into a low-dimensional space (i-vectors) rather than audio recordings. These changes made the challenge mo...
متن کاملAnalysis of i-vector framework for speaker identification in TV-shows
Inspired from the Joint Factor Analysis, the I-vector-based analysis has become the most popular and state-of-the-art framework for the speaker verification task. Mainly applied within the NIST/SRE evaluation campaigns, many studies have been proposed to improve more and more performance of speaker verification systems. Nevertheless, while the i-vector framework has been used in other speech pr...
متن کامل