Euclidean Distances as measures of speaker similarity including identical twin pairs: A forensic investigation using source and filter voice characteristics

Segundo, Eugenia San; Tsanas, Athanasios; Gómez-Vilda, Pedro

doi:10.1016/j.forsciint.2016.11.020

Cited by 38 publications

(34 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…However, few studies have attempted to address such issues ( [10,11] are exceptions). Notably, [12] evaluated the falsely accepted pairs (different-speaker classified as the same speaker) by an i-vector ASR system using auditory and acoustic phonetic analyses.…”

Section: 3! Integration Of Phonetic and Automatic Methodsmentioning

confidence: 99%

“…Although the differences between pairs resulted from a range of features, voice quality (VQ) was considered of Òfundamental importanceÓ for distinguishing voices. However, while the phonetic analysis highlighted the Ôback-endÕ value of VQ in separating speakers, it did not address how this information might be useful for front-end prediction of problematic speakers for the automatic system [11].…”

Section: 3! Integration Of Phonetic and Automatic Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Mapping Across Feature Spaces in Forensic Voice Comparison: The Contribution of Auditory-Based Voice Quality to (Semi-)Automatic System Testing

et al. 2017

View full text Add to dashboard Cite

In forensic voice comparison, there is increasing focus on the integration of automatic and phonetic methods to improve the validity and reliability of voice evidence to the courts. In line with this, we present a comparison of long-term measures of the speech signal to assess the extent to which they capture complementary speaker-specific information. Likelihood ratiobased testing was conducted using MFCCs and (linear and Melweighted) long-term formant distributions (LTFDs). Fusing automatic and semi-automatic systems yielded limited improvement in performance over the baseline MFCC system, indicating that these measures capture essentially the same speaker-specific information. The output from the best performing system was used to evaluate the contribution of auditory-based analysis of supralaryngeal (filter) and laryngeal (source) voice quality in system testing. Results suggest that the problematic speakers for the (semi-)automatic system are, to some extent, predictable from their supralaryngeal voice quality profiles, with the least distinctive speakers producing the weakest evidence and most misclassifications. However, the misclassified pairs were still easily differentiated via auditory analysis. Laryngeal voice quality may thus be useful in resolving problematic pairs for (semi-)automatic systems, potentially improving their overall performance.

show abstract

Section: 3! Integration Of Phonetic and Automatic Methodsmentioning

confidence: 99%

Section: 3! Integration Of Phonetic and Automatic Methodsmentioning

confidence: 99%

Mapping Across Feature Spaces in Forensic Voice Comparison: The Contribution of Auditory-Based Voice Quality to (Semi-)Automatic System Testing

et al. 2017

View full text Add to dashboard Cite

show abstract

“…The Euclidean distance was already successfully used as a similarity measure between VQ characterisations in forensic speaker comparison (San Segundo et al, 2017). The normalization of the mean posteriors by the posterior features from the modal or healthy speakers is conceptually similar to likelihood ratio test in speaker recognition (Hansen and Hasan, 2015), where likelihoods from the speaker model are subtracted by likelihoods obtained from the background/world model.…”

Section: Differential Phonological Posteriorsmentioning

confidence: 99%

Characterisation of voice quality of Parkinson’s disease using differential phonological posterior features

Cerňak

Orozco-Arroyave

Rudzicz

et al. 2017

Computer Speech & Language

View full text Add to dashboard Cite

Change in voice quality (VQ) is one of the first precursors of Parkinson's disease (PD). Specifically, impacted phonation and articulation causes the patient to have a breathy, husky-semiwhisper and hoarse voice.A goal of this paper is to characterize a VQ spectrum -the composition of non-modal phonations -of voice in PD. The paper relates non-modal healthy phonations: breathy, creaky, tense, falsetto and harsh, with disordered phonation in PD. First, statistics are learned to differentiate the modal and non-modal phonations. Statistics are computed using phonological posteriors, the probabilities of phonological features inferred from the speech signal using a deep learning approach. Second, statistics of disordered speech are learned from PD speech data comprising 50 patients and 50 healthy controls. Third, Euclidean distance is used to calculate similarity of non-modal and disordered statistics, and the inverse of the distances is used to obtain the composition of non-modal phonation in PD. Thus, pathological voice quality is characterised using healthy non-modal voice quality "base/eigenspace". The obtained results are interpreted as the voice of an average patient with PD and can be characterised by the voice quality spectrum composed of 30% breathy voice, 23% creaky voice, 20% tense voice, 15% falsetto voice and 12% harsh voice. In addition, the proposed features were applied for prediction * Corresponding author Email address: milos.cernak@idiap.ch (Milos Cernak) of the dysarthria level according to the Frenchay assessment score related to the larynx, and significant improvement is obtained for reading speech task. The proposed characterisation of VQ might also be applied to other kinds of pathological speech.

show abstract

“…Con el fin de mostrar el funcionamiento del sistema se lleva a cabo una prueba piloto con 11 usuarios diferentes, cada uno con 4 muestras de voz (considerando lo reportado en [8] donde se realiza una prueba con 12 usuarios). Para los procesos de medición y análisis de resultados, tanto por distancia como por puntaje, solo se tienen en cuenta los 5 primeros registros con mejor resultado.…”

Section: Resultsunclassified

“…Por su parte en [8], se señala que existe un crecimiento en los enfoques híbridos para una caracterización exitosa de la voz, particularmente en aplicaciones forenses, utilizando las características de la grabación de la señal como también de las obtenidas a través de filtros.…”

Section: Sistemas De Reconocimiento De Vozunclassified

Implementación de un algoritmo para la identificación de usuarios considerando problemas fisiológicos que afectan el habla

Rey-Lancheros

Gavilán-Acosta

Espitia

2017

ITECKNE Innovación E Investigación en Ingeniería

View full text Add to dashboard Cite

Euclidean Distances as measures of speaker similarity including identical twin pairs: A forensic investigation using source and filter voice characteristics

Cited by 38 publications

References 35 publications

Mapping Across Feature Spaces in Forensic Voice Comparison: The Contribution of Auditory-Based Voice Quality to (Semi-)Automatic System Testing

Mapping Across Feature Spaces in Forensic Voice Comparison: The Contribution of Auditory-Based Voice Quality to (Semi-)Automatic System Testing

Characterisation of voice quality of Parkinson’s disease using differential phonological posterior features

Implementación de un algoritmo para la identificación de usuarios considerando problemas fisiológicos que afectan el habla

Contact Info

Product

Resources

About