Year: 2025
Type: article
Abstract: The paper analyzes speaker voice identification methods that are widely used in the practice of flight accidents investigation by speech recorded in the cockpit of a moving aircraft during an emergency situation on board. The advantages and disadvantages of each of methods are described, the factors reducing the reliability of the identification are indicated, and the methodological foundations for using utterances of persons recorded on a phonogram, whose attribution has already been established as an additional data source, are formulated. The prerequisites for constructing a step-by-step methodology for attributing utterances of persons recorded in the cockpit of a moving aircraft using the latest achievements in the field of forensic examination of audio recordings are given.
Authors: Valery I. Arbuzov, Ivan S. Siparov
Institution: Saint Petersburg State University of Civil Aviation
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: bronze
OpenAlex ID: W4411519711