Homogeneity Measure Impact on Target and Non-Target Trials in Forensic Voice Comparison

It is common to see mobile recordings being presented as a forensic trace in a court. In such cases, a forensic expert is asked to analyze both suspect and criminal’s voice samples in order to determine the strength-of-evidence. This process is known as Forensic Voice Comparison (FVC). The Likelihood ratio (LR) framework is commonly used by the experts and quite often required by the expert’s associations “best practice guides”. Nevertheless, the LR accepts some practical limitations due both to intrinsic aspects of its estimation process and the information used during the FVC process. These aspects are embedded in a more general one, the lack of knowledge on FVC reliability. The question of reliability remains a major challenge, particularly for FVC systems where numerous variation factors like duration, noise, linguistic content or... within-speaker variability are not taken into account. Recently, we proposed an information theory-based criterion able to estimate one of these factors, the homogeneity of information between the two sides of a FVC trial. Thanks to this new criterion, we wish to explore new aspects of homogeneity in this article. We wish to question the impact of homogeneity on reliability separately on target and non-target trials. The study is performed using FABIOLE, a publicly available database dedicated to this kind of studies with a large number of recordings per target speaker. Our experiments report large differences of homogeneity impact between FVC genuine and impostor trials. These results show clearly the importance of intra-speaker variability effects in FVC reliability estimation. This study confirms also the interest of homogeneity measure for FVC reliability.