Karnataka - Mysore   

Voice analysis can help police nab criminals

— PHOTO: M.A. SRIRAM

Technology: S.K. Jain, assistant director, Central Forensic Science Laboratory, Chandigarh (centre) releasing a souvenir during the inauguration of a national workshop on ‘Forensic voice identification’ at the All-India Institute of Speech and Hearing (AIISH) in Mysore on Monday.

MYSORE: “Voice” has become significant in the investigation of criminal cases as most of the police officers use the latest technology of recording voices of suspects to narrow down suspects or eliminate the innocents during interrogation. The scientific analysis of the voice can further provide conclusive proof to link the suspect with the crime.

“Even though a lot of development has taken place in the area of speech, but much scientific development is still required to develop software and matching hardware that can analyse the subtle parameters of voice,” says S.K. Jain, assistant director, Central Forensic Science Laboratory, Chandigarh.

He was speaking after inaugurating a national workshop on “Forensic voice identification” at the All-India Institute of Speech and Hearing (AIISH) here on Monday.

Dr. Jain said forensic science laboratories had to deal with a large number of voice samples in the recent cases of match fixing, Parliament attack case and Tehelka tapes. “The laboratories get voice of any language, in any mode and with any environmental effect. It is, therefore, required to know about the acoustic, phonetic, speech enhancement, vocal anatomy and organ of speech and advanced computer approaches for analysis of audio and video samples,” he said.

Speaker-recognition was one of the most important emerging areas in forensic science as communication technology had reached a high level. Criminals had also been using telephones, mobile phones, wireless phones and tape recorded conversions in cases of kidnapping, threatening calls, anonymous calls, drug peddling and other nefarious activities.

“In these circumstances, the voice of individual is one of the most important clues to link the criminal with the crime or vice-versa, if recorded,” he added.

Dr Jain said: “The task of forensic speaker identification is highly complicated and tedious because it is essential to understand the variations in a speech. The variation could be due to intra–speaker variability (speaking in different languages, different sessions, effect of health such as cough, emotions); forced variability (mimic or disguise) and instrument variability (microphone, bandwidth).”

Some sources of these variations are transmission channels, recording devices, environmental noise and other physiological factors, he added.

In her address, director of AIISH Vijayalakshmi Basavaraj said several forensic laboratories in India had experts in fingerprints. However, they did not have training in speaker-identification. Therefore, the workshop had been organised to train them and speech pathologists in forensic voice identification. Thirty delegates from across the country are participating in the four-day workshop.

Lectures, demonstrations, discussions and practical sessions on several aspects of speaker identification will be held at the workshop. A group of experts from the AIISH and other institutions headed by S.R. Savithri, professor of Speech Sciences, Department of Speech-Language Sciences, is conducting the workshop.

Printer friendly page  
Send this article to Friends by E-Mail

Stories in this Section

Highlighting a message
Accused dies in Mysore jail
Highlighting a message
Voice analysis can help police nab criminals
International meet on food processing technology from tomorrow
Urbanisation taking its toll on ecosystem
Paragotsava from tomorrow
Girl falls into open water sump, dies

Copyright © 2007, The Hindu. Republication or redissemination of the contents of this screen are expressly prohibited without the written consent of The Hindu