Italian monozygotic twins’ speech: A preliminary forensic investigation

Alice Albanesi; Sonia Cenceschi; Chiara Meluzzi; Alessandro Trivilini

doi:10.17469/O2108AISV000008

Authors

Alice Albanesi Digital Forensic Service, Department of Innovative Technologies, University of Applied Sciences and Arts of Southern Switzerland, Via la Santa 1, 6962 Viganello, Switzerland https://orcid.org/0000-0003-0806-9048
Sonia Cenceschi Digital Forensic Service, Department of Innovative Technologies, University of Applied Sciences and Arts of Southern Switzerland, Via la Santa 1, 6962 Viganello, Switzerland https://orcid.org/0000-0002-4145-9593
Chiara Meluzzi Department of Literary Studies, Philology and Linguistics, Università degli Studi di Milano Statale, Via Festa del Perdono, 7 20122 Milan, Italy https://orcid.org/0000-0002-2291-006X
Alessandro Trivilini Digital Forensic Service, Department of Innovative Technologies, University of Applied Sciences and Arts of Southern Switzerland, Via la Santa 1, 6962 Viganello, Switzerland https://orcid.org/0000-0003-0687-801X

DOI:

https://doi.org/10.17469/O2108AISV000008

Keywords:

speaker recognition, monozygotic twins, normalization, formants, between-speaker variability

Abstract

In this study, we investigate whether it is possible to distinguish a speaker from his twin in low quality audio recordings. An analysis of both qualitative and quantitative data was conducted to compare the speech of 4 pairs of Italian twins (2F, 2M). The distributions of fundamental frequency and formants were similar across twin pairs, but Lobanov’s normalization allowed a differentiation of twins’ speech, especially in the elicited form. The statistical analysis confirmed these outcomes and highlighted some differences and the role of F3. The results are discussed in a forensic perspective. Further experiments will widen the sample and the features of interest to determine if this methodology represents a valid procedure for twins’ speech discrimination.