Problems and Prospects of Forensic Automatic Speaker Recognition by Voice
##plugins.themes.bootstrap3.article.main##
Abstract
Technologies of speaker recognition by voice in criminology were applied recently. Now they become topical in many countries.
Main methods, used in automatic speaker recognition by voice were considered in this article. Analysis of forensic speaker identification by voice and automatic identification system was performed. Outlook of novel automatic speaker recognition systems in forensic inspection, criminal search and strategical work were discussed. Results of automatic speaker identification evaluation, using database were described. Conclusion can be made, that automatic speaker recognition system is convenient, because it requires minimal knowledge from the user. This system can by applied in criminal search, but it has limitations in forensic inspection.
One of the most important lack of automatic speaker recognition systems is less accuracy of recognition and relatively long record length is required for training and testing of the system. In spite of lack of automatic recognition system, some methods that are used in this system can be applied in combined system. And how our and other countries experience shows, it allows to perform research rather and more objective. One of the most promising way could be development of the database, consisting of voice records and integrated automatic recognition system, what would allow to perform not only automatic speaker recognition, but finding test and template records too, and so on.
Main methods, used in automatic speaker recognition by voice were considered in this article. Analysis of forensic speaker identification by voice and automatic identification system was performed. Outlook of novel automatic speaker recognition systems in forensic inspection, criminal search and strategical work were discussed. Results of automatic speaker identification evaluation, using database were described. Conclusion can be made, that automatic speaker recognition system is convenient, because it requires minimal knowledge from the user. This system can by applied in criminal search, but it has limitations in forensic inspection.
One of the most important lack of automatic speaker recognition systems is less accuracy of recognition and relatively long record length is required for training and testing of the system. In spite of lack of automatic recognition system, some methods that are used in this system can be applied in combined system. And how our and other countries experience shows, it allows to perform research rather and more objective. One of the most promising way could be development of the database, consisting of voice records and integrated automatic recognition system, what would allow to perform not only automatic speaker recognition, but finding test and template records too, and so on.
##plugins.themes.bootstrap3.article.details##
Section
Articles
Authors contributing to Jurisprudence agree to publish their articles under a Creative Commons Attribution-NoDerivatives 4.0 International Public (CC BY-NC-ND) License, allowing third parties to share their work (copy, distribute, transmit) and to adapt it, under the condition that the authors are given credit, and that in the event of reuse or distribution, the terms of this licence are made clear.
Authors retain copyright of their work, with first publication rights granted to the Association for Learning Technology.
Please see Copyright and Licence Agreement for further details.
Authors retain copyright of their work, with first publication rights granted to the Association for Learning Technology.
Please see Copyright and Licence Agreement for further details.
