The comparison of “Idiot's Bayes” and multivariate kernel-density in forensic speaker identification using Chinese vowel /a/

摘要: The discriminant performance of likelihood ratios based on “Idiot's Bayes” approach and multivariate kernel density were examined the speech 21 male Chinese speakers using telephone recordings. parameters used in this paper are formant central-frequencies, extracted from vowel /a/. Experimental results show that provides a little stronger support for same-speaker hypothesis much different-speaker hypothesis; they both have good