Fig 8.TIF (1.26 MB)
Text-dependent speaker identification performance of the proposed and existing methods using the UM database.
figure
posted on 2016-07-08, 16:59 authored by Md. Atiqul Islam, Wissam A. Jassim, Ng Siew Cheok, Muhammad Shamsul Arefeen ZilanyTwo cases of frequency bands were considered: left panels show the performance of the SI systems using features from the narrowband frequencies (<1 kHz), and the right panels represent performances for the wideband frequencies. Results are shown as a function of SNR with three different types of noise (A: white Gaussian noise, B: pink noise, and C: street noise). Speech samples from 39 speakers were used for evaluation and comparison of the performance of different methods.
History
Usage metrics
Categories
Keywords
2- D neurogramsGamma-tone frequency cepstral coefficientsMel-frequency cepstral coefficientsAuditory Periphery Speaker identificationspeaker identification systemspeech processing applicationsRobust Speaker Identification Systemspeaker identification methodsGaussian mixture model-universal background model classification techniqueperformanceresponsespeech signalstext-dependent speaker databases
Licence
Exports
RefWorks
BibTeX
Ref. manager
Endnote
DataCite
NLM
DC