Forensic Application of Speaker Identification
نویسندگان
چکیده
This paper describes a forensic application scenario in the speaker recognition domain and explains the techniques required to develop a complete forensic analysis system based on speaker identification. It is presented a standard manner of how Gaussian Mixture Models (GMMs) could be used in order to compute and express identification results in a judicial system. A review of possible techniques in forensic expertise is made. Experiments on a filtered POTS telephone database are exemplified from the forensics examiner perspective.
منابع مشابه
The Application of a Bayesian Approach to Auditory Analysis in Forensic Speaker Identification
Research into auditory analysis in forensic speaker identification has attracted little attention, yet it is regarded as an essential component of forensic speaker identification. One possible reason for this lack of interest may be associated with the difficulty of knowing how to present the results of auditory analysis in a logical and quantifiable way. The use of likelihood ratios enables th...
متن کاملTechnical forensic speaker identification from a Bayesian linguist's perspective
Important methodological aspects of Technical Forensic Speaker Identification are discussed and exemplified. The centrality of the Likelihood Ratio of Bayes' Theorem as the proper way of forensically evaluating speech evidence is emphasised, as well as the many different types of evidence that are of use in discriminating same-speaker from differentspeaker speech samples.
متن کاملPhonetic Analysis in Forensic Speaker Identification. an Example of Routine Expert Actions.
Forensic speaker identification is one of the most important tasks dealt with in forensic phonetics. The usage of acoustic evidence increases and at the same time grows the necessity for easy-toapply and reliable methods. The method should be consistent and straight and should enable an expert to make identification decision with clearly determined probability. This paper is to outline the rout...
متن کاملChannel Compensation for Forensic Speaker Identification Using Inverse Processing
Typically, speaker identification examination requires two audio recordings: a voice sample and a questionable recording. The questionable one is in most of the cases the intercepted or recorded phone call. As mobile phones became the most popular way of communication, the largest number of questionable recordings comes from GSM channels. They use special algorithms and devices to transmit the ...
متن کاملComparison of speaker recognition systems on a real forensic benchmark
This paper analyses the performance of several automatic speaker recognition systems using a real forensic database. The systems evaluated have been tested or are currently in use by forensic institutes. A comprehensive error analysis is performed in order to assess the each system’s behaviour to real casework. We further investigate compensation techniques aimed at minimising the performance g...
متن کامل