Title: Aural-acoustic analysis and gender identification of morphed male and female audios

Authors: Palak Aneja; Sumit Kumar Choudhary; Surbhi Mathur

Addresses: School of Forensic Science, Risk Management and National Security, Rashtriya Raksha University, Gandhinagar, Gujarat, India ' School of Forensic Science, Risk Management and National Security, Rashtriya Raksha University, Gandhinagar, Gujarat, India ' School of Forensic Science, National Forensic Science University, Gandhinagar, Gujarat, India

Abstract: Hiding the identity of the criminal is essential from their perspective to avoid getting caught. Crimes like kidnapping, threat calls, and ransom calls often involve the voice of criminals as a crucial piece of evidence. The person's voice is used in the biometric systems for identification as it is unique. Discriminating gender from the questioned audios helps shorten the list of suspects in any offence. In this paper, male voices were compared with morphed male audio and female voices were compared with morphed female audio. To disguise the identity and to study the reliability of the various speaker identification parameters, the female audios were morphed into male audios; and the male audios were morphed into female audios using the same morphing software. In this experiment, 50 female and 50 male voices samples were converted using morphing software, and auditory and acoustic analysis was done for forensic speaker identification. The aural parameters like speech rate, articulation, delivery of speech, dynamic loudness and acoustic parameters like pitch and formant frequency for five vowels were compared.

Keywords: audio; male; female; morphing; speaker; identification; disguise.

DOI: 10.1504/IJESDF.2023.129275

International Journal of Electronic Security and Digital Forensics, 2023 Vol.15 No.2, pp.177 - 183

Received: 03 Mar 2022
Accepted: 22 Jun 2022
Published online: 02 Mar 2023 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article