Empirical validation of likelihood ratio methods: A case study in forensic speaker recognition

Filipe Guapo; Paulo Correia; Didier Meuwly; David Van Der Vloed

doi:10.1109/IWBF.2016.7449678

Empirical validation of likelihood ratio methods: A case study in forensic speaker recognition

Filipe Guapo, Paulo Correia, Didier Meuwly, David Van Der Vloed

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

4 Citations (Scopus)

Abstract

In recent years there has been an increase in the offer of commercial software tools for forensic evaluation of biometric evidence. The important role these tools are meant to play requires that these are subject to a rigorous and careful validation process. This has spurred guidelines and procedures through which the validation should be processed. In this work we describe the validation procedure of a commercial Likelihood Ratio computation tool while proposing a new approach to performance evaluation. This is done by testing not only the tool alongside a single baseline method but several baseline methods which have previously been validated by the community. The tool is tested with over 90000 speaker recognition scores and its performance is analyzed and compared to well-known methods, it is shown that the use of multiple baseline methods increases the amount of information and context. This improves decision making in the validation process and helps safeguarding against outliers.

Original language	English
Title of host publication	Proceedings 2016 4th International Conference on Biometrics and Forensics (IWBF)
Publisher	IEEE
ISBN (Electronic)	978-1-4673-9448-2
DOIs	https://doi.org/10.1109/IWBF.2016.7449678
Publication status	Published - 9 Apr 2016
Externally published	Yes
Event	4th International Workshop on Biometrics and Forensics, IWBF 2016 - Limassol, Cyprus Duration: 3 Mar 2016 → 4 Mar 2016 Conference number: 4

Publication series

Name	Proceedings - 2016 4th International Workshop on Biometrics and Forensics, IWBF 2016

Conference

Conference	4th International Workshop on Biometrics and Forensics, IWBF 2016
Abbreviated title	IWBF 2016
Country/Territory	Cyprus
City	Limassol
Period	3/03/16 → 4/03/16

Keywords

n/a OA procedure
Forensic evaluation
Forensic speaker recognition
Forensic tool validation
Likelihood ratio
Forensic biometrics

Access to Document

10.1109/IWBF.2016.7449678

Cite this

Guapo, F., Correia, P., Meuwly, D., & Van Der Vloed, D. (2016). Empirical validation of likelihood ratio methods: A case study in forensic speaker recognition. In Proceedings 2016 4th International Conference on Biometrics and Forensics (IWBF) Article 7449678 (Proceedings - 2016 4th International Workshop on Biometrics and Forensics, IWBF 2016). IEEE. https://doi.org/10.1109/IWBF.2016.7449678

@inproceedings{400680db41d146e18f220918b6f60aba,

title = "Empirical validation of likelihood ratio methods: A case study in forensic speaker recognition",

abstract = "In recent years there has been an increase in the offer of commercial software tools for forensic evaluation of biometric evidence. The important role these tools are meant to play requires that these are subject to a rigorous and careful validation process. This has spurred guidelines and procedures through which the validation should be processed. In this work we describe the validation procedure of a commercial Likelihood Ratio computation tool while proposing a new approach to performance evaluation. This is done by testing not only the tool alongside a single baseline method but several baseline methods which have previously been validated by the community. The tool is tested with over 90000 speaker recognition scores and its performance is analyzed and compared to well-known methods, it is shown that the use of multiple baseline methods increases the amount of information and context. This improves decision making in the validation process and helps safeguarding against outliers.",

keywords = "n/a OA procedure, Forensic evaluation, Forensic speaker recognition, Forensic tool validation, Likelihood ratio, Forensic biometrics",

author = "Filipe Guapo and Paulo Correia and Didier Meuwly and {Van Der Vloed}, David",

note = "Publisher Copyright: {\textcopyright} 2016 IEEE.; 4th International Workshop on Biometrics and Forensics, IWBF 2016, IWBF 2016 ; Conference date: 03-03-2016 Through 04-03-2016",

year = "2016",

month = apr,

day = "9",

doi = "10.1109/IWBF.2016.7449678",

language = "English",

series = "Proceedings - 2016 4th International Workshop on Biometrics and Forensics, IWBF 2016",

publisher = "IEEE",

booktitle = "Proceedings 2016 4th International Conference on Biometrics and Forensics (IWBF)",

address = "United States",

}

Guapo, F, Correia, P, Meuwly, D & Van Der Vloed, D 2016, Empirical validation of likelihood ratio methods: A case study in forensic speaker recognition. in Proceedings 2016 4th International Conference on Biometrics and Forensics (IWBF)., 7449678, Proceedings - 2016 4th International Workshop on Biometrics and Forensics, IWBF 2016, IEEE, 4th International Workshop on Biometrics and Forensics, IWBF 2016, Limassol, Cyprus, 3/03/16. https://doi.org/10.1109/IWBF.2016.7449678

Empirical validation of likelihood ratio methods: A case study in forensic speaker recognition. / Guapo, Filipe; Correia, Paulo; Meuwly, Didier et al.
Proceedings 2016 4th International Conference on Biometrics and Forensics (IWBF). IEEE, 2016. 7449678 (Proceedings - 2016 4th International Workshop on Biometrics and Forensics, IWBF 2016).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

TY - GEN

T1 - Empirical validation of likelihood ratio methods

T2 - 4th International Workshop on Biometrics and Forensics, IWBF 2016

AU - Guapo, Filipe

AU - Correia, Paulo

AU - Meuwly, Didier

AU - Van Der Vloed, David

N1 - Conference code: 4

PY - 2016/4/9

Y1 - 2016/4/9

N2 - In recent years there has been an increase in the offer of commercial software tools for forensic evaluation of biometric evidence. The important role these tools are meant to play requires that these are subject to a rigorous and careful validation process. This has spurred guidelines and procedures through which the validation should be processed. In this work we describe the validation procedure of a commercial Likelihood Ratio computation tool while proposing a new approach to performance evaluation. This is done by testing not only the tool alongside a single baseline method but several baseline methods which have previously been validated by the community. The tool is tested with over 90000 speaker recognition scores and its performance is analyzed and compared to well-known methods, it is shown that the use of multiple baseline methods increases the amount of information and context. This improves decision making in the validation process and helps safeguarding against outliers.

AB - In recent years there has been an increase in the offer of commercial software tools for forensic evaluation of biometric evidence. The important role these tools are meant to play requires that these are subject to a rigorous and careful validation process. This has spurred guidelines and procedures through which the validation should be processed. In this work we describe the validation procedure of a commercial Likelihood Ratio computation tool while proposing a new approach to performance evaluation. This is done by testing not only the tool alongside a single baseline method but several baseline methods which have previously been validated by the community. The tool is tested with over 90000 speaker recognition scores and its performance is analyzed and compared to well-known methods, it is shown that the use of multiple baseline methods increases the amount of information and context. This improves decision making in the validation process and helps safeguarding against outliers.

KW - n/a OA procedure

KW - Forensic evaluation

KW - Forensic speaker recognition

KW - Forensic tool validation

KW - Likelihood ratio

KW - Forensic biometrics

U2 - 10.1109/IWBF.2016.7449678

DO - 10.1109/IWBF.2016.7449678

M3 - Conference contribution

AN - SCOPUS:84965147652

T3 - Proceedings - 2016 4th International Workshop on Biometrics and Forensics, IWBF 2016

BT - Proceedings 2016 4th International Conference on Biometrics and Forensics (IWBF)

PB - IEEE

Y2 - 3 March 2016 through 4 March 2016

ER -

Empirical validation of likelihood ratio methods: A case study in forensic speaker recognition

Abstract

Publication series

Conference

Keywords

Access to Document

Fingerprint

Cite this