PROBLEM OF SPEAKER IDENTIFICATION AND ELIMINATION (FROM SELECTED MATERIALS ON LEGAL AND TECHNICAL ASPECTS OF 'VOICEPRINT' IDENTIFICATION, 1976 - SEE NCJ-44202)

• NCJ Number: 44203
• Author(s): O I TOSI
• Date Published: 1976
• Length: 33 pages
• Corporate Author: National College of Criminal Defense Lawyers and Public Defenders
• Publication Format: Document
• Language: English
• Country: United States of America

Annotation

THEORIES OF SPEECH PRODUCTION AND SPEAKER VARIABILITY ARE REVIEWED, METHODS OF SPEAKER IDENTIFICATION AND ELIMINATION ARE DESCRIBED, AND THE LEGAL STATUS OF SPEAKER IDENTIFICATION EVIDENCE IS ASSESSED.

Abstract

VOICE IDENTIFICATION AND ELIMINATION METHODS FALL ALONG A SUBJECTIVE-OBJECTIVE CONTINUUM, THEIR POSITION ON THE CONTINUUM DEPENDING ON THE RELATIVE INVOLVEMENT OF THE HUMAN MIND AND OF MECHANICAL OR ELECTRONIC MEANS. EXPERIMENTAL DATA ON AURAL EXAMINATION OF VOICES AND ON VISUAL EXAMINATION OF SPECTROGRAMS (SPECTROGRAPHY, VOICEPRINT IDENTIFICATION) ARE SUMMARIZED, AS ARE DATA ON AUTOMATIC AND SEMIAUTOMATIC OBJECTIVE METHODS OF SPEAKER RECOGNITION. IT IS NOTED THAT THE OBJECTIVE METHODS ARE IN THE EARLY STAGES OF DEVELOPMENT AND AS YET ARE MORE SUITABLE FOR SPEAKER AUTHENTICATION OR VERIFICATION THAN FOR SPEAKER IDENTIFICATION IN LEGAL CASES. COURT CASES RELEVANT TO THE USE OF VOICE IDENTIFICATION EVIDENCE PRIOR TO AND AFTER 1970 ARE REVIEWED, AS ARE THE ACTIVITIES OF THE VOICE IDENTIFICATION UNIT OF THE MICHIGAN DEPARTMENT OF STATE POLICE AND THE INTERNATIONAL ASSOCIATION OF VOICE IDENTIFICATION. IT IS CONCLUDED THAT NO COMBINATION OF METHODS MAY EVER PRODUCE ABSOLUTELY POSITIVE VOICE IDENTIFICATION OR ELIMINATION IN ALL CASES. THE BETTER THE QUALITY OF AVAILABLE VOICE SAMPLES, THE BETTER THE QUALIFICATIONS OF THE EXAMINER, AND THE MORE COMPREHENSIVE THE CLUSTER OF METHODS USED, THE BETTER THE CHANCE OF OBTAINING RELIABLE IDENTIFICATION. CONTINUING RESEARCH ON SUBJECTIVE AND OBJECTIVE METHODS OF VOICE IDENTIFICATION IS URGED. A LIST OF REFERENCES IS INCLUDED.