Dr. Mia Šešum & Dr. Ema Petrović,
Department of Special Education and Rehabilitation of the Deaf and Hard of Hearing Persons, Faculty of Special Education and Rehabilitation, University of Belgrade, Belgrade, Serbia & Department of Oriental Studies, Faculty of Philology, University of Belgrade, Belgrade, Serbia
The voice and speech of every human being are determined by their anatomical and physiological characteristics, as well as by the speech habits acquired during life. From a forensic point of view, voice and speech are inextricably linked, as voice production is a necessary precondition for speech production. The theory of speaker identification is based on the premise that, given the complexity of the speech system and learned speech behavior, no two people can have an identical voice and speech. Since voice and speech are mutable on conscious and unconscious level, concerning both internal and external factors, and are also traceable through the temporal dimension, speaker identification is considered the most complex discipline within forensic science. The article systematically presents the methods used in the past and those used today to identify speakers, focusing in particular on their advantages and disadvantages. The literature search was carried out in the databases of the Serbian Library Consortium and in relevant internet search engines. The literature review has shown that the rapid development of technology in the 21st century has significantly influenced forensic phonetics and increased the need for its application. However, this has not improved the reliability of the results, as the complexity of speech expression cannot yet be solved by automated speaker identification. Therefore, it is necessary to carefully select and train experts working in this field, in order to ensure reliable results relevant to the needs of court proceedings.
Keywords: Speaker Identification, Auditive-Instrumental Method, Voice and Speech, Spectography, Automatic Method
The above abstract is a part of the article which was accepted at The Ninth International Conference on Languages, Linguistics, Translation and Literature (WWW.LLLD.IR), 1-2 February 2024, Ahwaz.
