A statistical procedure to adjust for time-interval mismatch in forensic voice comparison

Geoffrey Stewart Morrison; F. Kelly

doi:10.1016/j.specom.2019.07.001

A statistical procedure to adjust for time-interval mismatch in forensic voice comparison

Research output: Contribution to journal › Article › peer-review

Abstract

The present paper describes a statistical modeling procedure that was developed to account for the fact that, in a forensic voice comparison analysis conducted for a particular case, there was a long time interval between when the questioned- and known-speaker recordings were made (six years), but in the sample of the relevant population used for training and testing the forensic voice comparison system there was a short interval (hours to days) between when each of multiple recordings of each speaker was made. The present paper also includes results of empirical validation of the procedure. Although based on a particular case, the procedure has potential for wider application given that
relatively long time intervals between the recording of questioned and known speakers are not uncommon in casework.

Original language	English
Pages (from-to)	15-21
Number of pages	7
Journal	Speech Communication
Volume	112
Early online date	2 Jul 2019
DOIs	https://doi.org/10.1016/j.specom.2019.07.001
Publication status	Published - 1 Sept 2019

Bibliographical note

Keywords

Forensic voice comparison
Likelihood ratio
Mismatch
Time interval

Access to Document

10.1016/j.specom.2019.07.001Licence: CC BY 3.0

A statistical procedure to adjust for time-interval mismatch in forensic voice comparisonFinal published version, 1.38 MBLicence: CC BY 3.0

Cite this

@article{b18ec537f9e74fec8a1d77dc8acd9fe9,

title = "A statistical procedure to adjust for time-interval mismatch in forensic voice comparison",

abstract = "The present paper describes a statistical modeling procedure that was developed to account for the fact that, in a forensic voice comparison analysis conducted for a particular case, there was a long time interval between when the questioned- and known-speaker recordings were made (six years), but in the sample of the relevant population used for training and testing the forensic voice comparison system there was a short interval (hours to days) between when each of multiple recordings of each speaker was made. The present paper also includes results of empirical validation of the procedure. Although based on a particular case, the procedure has potential for wider application given that relatively long time intervals between the recording of questioned and known speakers are not uncommon in casework.",

keywords = "Forensic voice comparison, Likelihood ratio, Mismatch, Time interval",

author = "Morrison, {Geoffrey Stewart} and F. Kelly",

note = "{\textcopyright}2019 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license.(http://creativecommons.org/licenses/by/4.0/) ",

year = "2019",

month = sep,

day = "1",

doi = "10.1016/j.specom.2019.07.001",

language = "English",

volume = "112",

pages = "15--21",

journal = "Speech Communication",

issn = "0167-6393",

publisher = "Elsevier",

}

TY - JOUR

T1 - A statistical procedure to adjust for time-interval mismatch in forensic voice comparison

AU - Morrison, Geoffrey Stewart

AU - Kelly, F.

PY - 2019/9/1

Y1 - 2019/9/1

N2 - The present paper describes a statistical modeling procedure that was developed to account for the fact that, in a forensic voice comparison analysis conducted for a particular case, there was a long time interval between when the questioned- and known-speaker recordings were made (six years), but in the sample of the relevant population used for training and testing the forensic voice comparison system there was a short interval (hours to days) between when each of multiple recordings of each speaker was made. The present paper also includes results of empirical validation of the procedure. Although based on a particular case, the procedure has potential for wider application given that relatively long time intervals between the recording of questioned and known speakers are not uncommon in casework.

AB - The present paper describes a statistical modeling procedure that was developed to account for the fact that, in a forensic voice comparison analysis conducted for a particular case, there was a long time interval between when the questioned- and known-speaker recordings were made (six years), but in the sample of the relevant population used for training and testing the forensic voice comparison system there was a short interval (hours to days) between when each of multiple recordings of each speaker was made. The present paper also includes results of empirical validation of the procedure. Although based on a particular case, the procedure has potential for wider application given that relatively long time intervals between the recording of questioned and known speakers are not uncommon in casework.

KW - Forensic voice comparison

KW - Likelihood ratio

KW - Mismatch

KW - Time interval

UR - https://www.sciencedirect.com/science/article/pii/S0167639319300858?dgcid=raven_sd_search_email

UR - http://www.scopus.com/inward/record.url?scp=85068465338&partnerID=8YFLogxK

U2 - 10.1016/j.specom.2019.07.001

DO - 10.1016/j.specom.2019.07.001

M3 - Article

SN - 0167-6393

VL - 112

SP - 15

EP - 21

JO - Speech Communication

JF - Speech Communication

ER -

A statistical procedure to adjust for time-interval mismatch in forensic voice comparison

Abstract

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint