Forensic Voice Comparison: Human-Supervised-Automatic Approach

Morrison, Geoffrey Stewart; Weber, Philip; Enzinger, Ewald; Labrador, Beltrán; Lozano-Díez, Alicia; Ramos, Daniel; González-Rodríguez, Joaquín

doi:10.1016/b978-0-12-823677-2.00182-3

Cited by 6 publications

(2 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As a matter of example, the Euclidean distance, the Pearson correlation distance, or uncalibrated LRs from a feature-based LR model are already used as scores in different forensic disciplines. The good calibration properties of the SLRs make them used in different branches of forensic science such as glass comparison [6], MDMA comparison [9], forensic speaker recognition [10,11], or biometric systems [12,13]. However, in practice, their main disadvantage is that their performance relies on the selection of the score.…”

Section: Introductionmentioning

confidence: 99%

Gaussianization of LA-ICP-MS Features to Improve Calibration in Forensic Glass Comparison

et al. 2022

View full text Add to dashboard Cite

Section: Introductionmentioning

confidence: 99%

Gaussianization of LA-ICP-MS Features to Improve Calibration in Forensic Glass Comparison

et al. 2022

View full text Add to dashboard Cite

“…The score-based approaches typically present good calibration properties despite the scarcity of the training data [18,19]. These score-based LR methods have been successfully applied in many other forensic disciplines [20,21,22,23,24,25,15,14].…”

Section: Introductionmentioning

confidence: 99%

Improving calibration of forensic glass comparisons by considering uncertainty in feature-based elemental data

Ramos

Maroñas

Almirall

2021

Chemometrics and Intelligent Laboratory Systems

View full text Add to dashboard Cite

Effect of identical twins on deep speaker embeddings based forensic voice comparison

Abed,

Sztahó

2024

Int J Speech Technol

View full text Add to dashboard Cite

Deep learning has gained widespread adoption in forensic voice comparison in recent years. It is mainly used to learn speaker representations, known as embedding features or vectors. In this work, the effect of identical twins on two state-of-the-art deep speaker embedding methods was investigated with special focus on metrics of forensic voice comparison. The speaker verification performance has been assessed using the likelihood-ratio framework by likelihood ratio cost and equal error rate. The AVTD twin speech dataset was applied. The results show a significant reduction in speaker verification performance when twin samples are present. Neither the adaptation of LR score calculation to twin samples, nor fine-tuning the pre-trained speaker embedding models seemed to be able to leverage this limitation. It was found that the recognition of same or different speakers was possible even in the case of identical twins but the performance dropped greatly. The lowest EER of the best performing model was 3.4% in the case of non-twin; at the same time, EER was 25.3% when twins were present. This doesn’t mean that the presented methods are useless in case of identical twins, but it must be taken into consideration that in case of a higher likelihood-ratio score (which indicates same speakers on the tested samples), the possibility of twins must also be considered in a real casework.

show abstract

Forensic Voice Comparison: Human-Supervised-Automatic Approach

Cited by 6 publications

References 41 publications

Gaussianization of LA-ICP-MS Features to Improve Calibration in Forensic Glass Comparison

Gaussianization of LA-ICP-MS Features to Improve Calibration in Forensic Glass Comparison

Improving calibration of forensic glass comparisons by considering uncertainty in feature-based elemental data

Effect of identical twins on deep speaker embeddings based forensic voice comparison

Contact Info

Product

Resources

About