Crimino-Vision Events

A Webinar on the topic - "Forensic Speaker Identification" was organised by the UG Department of Criminology on May 8th, 2021 at 2:00 pm on Google meet platform. Mr. Arjun Shivakumar, Forensic Tutor, Department of Forensic Medicine & Toxicology,KMC Manipal was the resource person for this session.

He started the session by explaining the exact terminology of voice analysis that is speaker identification.

The ability of recognizing a person from his voice is called speaker recognition. He explained that speaker recognition can be divided into speaker identification where it is found if the voice belongs to the concerned group or person and speaker verification where the speaker is found.

The resource person covered various areas of forensic speaker identification and gave a thorough walk through of the procedures involved in identifying speakers of recordings along with methods used.

He explained the main methods used in speaker recognition which are Aural perceptual method, Visual Communication method and machines. In aural perceptual method the person tries to match the voice to an individual by paying attention to certain characteristics and listening to sounds in the environment that could help they perceive the audio through certain parameters like pitch, rate of speech etc. In visual inspection of spectrograms, two sets of spectrograms are compared. Later on voice print identification was developed. The resource person showed the students pictures of voice prints in spectrogram and explained how nasal sound is also shown in them. In semi automatic speaker identification the examiner selects known and unknown samples which are then processed by software.

Overall it was a very informative webinar. Students learnt how voice print analysis can be useful in many cases and how it's done. They got an insight into the various career opportunities in this field and got sufficient knowledge of how important voice recordings can be as evidence in the court of law.