Effects of the Phonological Contents and Transmission Channels on Forensic Speaker Recognition

Amino, Kanae; Osanai, Takashi; Kamada, Toshiaki; Makinae, Hisanori; Arai, Takayuki

doi:10.1007/978-1-4614-0263-3_11

Cited by 14 publications

(12 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The results are quite consistent between the 30 target speakers, with limited variations. This outcome corroborates results of [25,24,26,29], where oral vowels and nasals are found to be particularly speaker specific information. (non-target trials) per speaker and for "all".…”

Section: Resultssupporting

confidence: 90%

“…The influence of the phonological content of both voice recordings was also evaluated in [22] in which authors suggest that glides and liquids together, vowels -and more particularly nasal vowels-and nasal consonants contain more speaker-specific information than phonemically balanced speech utterances. According to [25,26,24,29], nasals and vowels were found to be particularly speaker specific information and nasal vowels are more discriminant than oral vowels. Finally, [23] and, more recently, [32], show that some frequency sub-bands seem to be more relevant to characterize speakers than some others.…”

Section: A Review Of Literaturementioning

confidence: 99%

“…Several research works like [22,23,24] agree that speaker specific information is not equally distributed on the speech signal and particularly depends on the phoneme distribution. For example, in [25,24,26] authors show that nasal and oral vowel convey the most important speaker specific information than the other phonemic categories. However, for ASR system based on I-vector, a recording is encoded by one low dimensional vector and thereby the phonological content of a recording is not used explicitly, as well as the presence or absence of different speaker-specific cues, which could be a key in some forensic cases.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Phonological content impact on wrongful convictions in Forensic Voice Comparison context

Ajili

Bonastre

Kheder

et al. 2017

2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

Forensic Voice Comparison (FVC) is increasingly using the likelihood ratio (LR) in order to indicate whether the evidence supports the prosecution (same-speaker) or defender (different-speakers) hypotheses. Nevertheless, the LR accepts some practical limitations due both to its estimation process itself and to a lack of knowledge about the reliability of this (practical) estimation process. It is particularly true when FVC is considered using Automatic Speaker Recognition (ASR) systems. Indeed, in the LR estimation performed by ASR systems, different factors are not considered such as speaker intrinsic characteristics, denoted "speaker factor", the amount of information involved in the comparison as well as the phonological content and so on. This article focuses on the impact of phonological content on FVC involving two different speakers and more precisely the potential implication of a specific phonemic category on wrongful conviction cases (innocents are send behind bars). We show that even though the vast majority of speaker pairs (more than 90%) are well discriminated, few pairs are difficult to distinguish. For the "best" discriminated pairs, all the phonemic content play a positive role in speaker discrimination while for the "worst" pairs, it appears that nasals have a negative effect and lead to a confusion between speakers.

show abstract

Section: Resultssupporting

confidence: 90%

Section: A Review Of Literaturementioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Phonological content impact on wrongful convictions in Forensic Voice Comparison context

Ajili

Bonastre

Kheder

et al. 2017

2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

show abstract

“…2 Contenu phonétique et caractérisation du locuteur L'étude de l'information spécifique du locuteur portées par les phonèmes individuels ou encore des classes de phonèmes a fait l'objet de différents travaux comme (Wolf, 1972;Sambur, 1975;Eatock & Mason, 1994;Hofker, 1977;Kashyap, 1976;Amino et al, 2006Amino et al, , 2012Antal & Toderean, 2006). Les voyelles orales et les nasales apparaissent en tête en termes de discrimination entre les locuteurs.…”

Section: Introductionunclassified

Comparaison des voix dans le cadre judiciaire : influence du contenu phonétique

Ajili¹,

Bonastre²,

Kheder³

et al. 2018

XXXIIe Journées D’Études Sur La Parole

View full text Add to dashboard Cite

En comparaison de voix dans le domaine criminalistique, l'approche Bayésienne est devenue le nouveau "golden standard". Dans cette approche, l'expert exprime ses résultats par un unique nombre, le rapport de vraisemblance (LR). Cet article s'intéresse à l'influence du contenu phonétique sur la fiabilité du LR. Nous nous intéressons particulièrement à la quantité d'information spécifique au locuteur que portent les différents sons de la parole. Cette étude met en évidence des différences importantes entre les phonèmes et, surtout, la forte influence de la variabilité intra-locuteur. ABSTRACT phonetic content impact on forensic voice comparison. Forensic Voice Comparison (FVC) is increasingly using the likelihood ratio (LR).This article focuses on the impact of phonemic content on FVC performance and variability. The results demonstrate the importance of the phonemic content and highlight interesting differences between inter-speakers effects and intra-speaker's ones.

show abstract

“…Front stressed vowels [7] and nasalized vowels [8] have been found to be particularly useful. The works in [9,10] have evinced that nasal sounds facilitate higher human speaker recognition than other consonants. This is attributable to the fact that the resonance cavities shaping nasal sounds differ considerably among speakers [11].…”

Section: Introductionmentioning

confidence: 99%

Importance of intelligible phonemes for human speaker recognition in different channel bandwidths

Gallardo¹,

Möller²,

Wagner³

2015

Interspeech 2015

View full text Add to dashboard Cite

It is known that nasal consonants and vowels are more effective than other phonemes for human speaker recognition. However, the influence of channel transmissions on the speakerdiscriminative capabilities of phonemes has not yet been examined. Specifically, the speech bandwidth has a strong effect on the human speaker recognition performance and also on the speech intelligibility. The phonemes that permit more accurate human speaker recognition are determined in this study by means of a speaker verification auditory test, focusing on the differences in performance when the stimuli are presented to listeners in narrowband and in wideband. The speech intelligibility is also investigated via an intelligibility test employing the same speech stimuli. Finally, the possible relationship between phonemes offering better human speaker recognition and more intelligible phonemes in the transition to an enhanced bandwidth is discussed.

show abstract

Effects of the Phonological Contents and Transmission Channels on Forensic Speaker Recognition

Cited by 14 publications

References 35 publications

Phonological content impact on wrongful convictions in Forensic Voice Comparison context

Phonological content impact on wrongful convictions in Forensic Voice Comparison context

Comparaison des voix dans le cadre judiciaire : influence du contenu phonétique

Importance of intelligible phonemes for human speaker recognition in different channel bandwidths

Contact Info

Product

Resources

About