The Intrinsic Forensic Discriminatory Power of Diphthongs

This paper describes an experiment investigating how well same-speaker speech samples can be discriminated from different-speaker speech samples using acoustic parameters from Australian English diphthongs. A two-level kernel density multivariate likelihood ratio is used as a discriminant function on five of the diphthongs of the 171 speakers of the Bernard data base: /a    /. Comparing samples with just one token per diphthong each, an EER of ca.10% is obtained. It is concluded that diphthongal F-pattern can yield useful strength of forensic speaker identification evidence.