Abstract
Speaker identification systems are based on the fact that speakers differ in their anatomy and in their speech habits. In identical twins the anatomy is identical, or at least very similar, so that speech habits are the main cause for speech differences. Ten pairs of identical twins were recorded and identified using a minimum distance classifier. The distribution of distances from a subject to his twin and to the "world" are analyzed using various features. F-Ratios and text independent identification results are presented and discussed.
| Original language | English |
|---|---|
| Pages | 213-216 |
| Number of pages | 4 |
| State | Published - 1 Jan 2019 |
| Event | ESCA Workshop on Automatic Speaker Recognition, Identification, and Verification, ASRIV 1994 - Martigny, Switzerland Duration: 7 Apr 1994 → 9 Apr 1994 |
Conference
| Conference | ESCA Workshop on Automatic Speaker Recognition, Identification, and Verification, ASRIV 1994 |
|---|---|
| Country/Territory | Switzerland |
| City | Martigny |
| Period | 7/04/94 → 9/04/94 |
Keywords
- Identical Twins
- Speaker Identification
ASJC Scopus subject areas
- Signal Processing
- Software
- Human-Computer Interaction