Hierarchical speaker identification using speaker clustering
Authors:
- Qiuhai ZhongQZQiuhai Zhong
- 8 July 2004
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 299-304
Abstract
We explore an approach to speaker identification called speaker clustering in the GMM-based speaker recognition system in order to reduce the computational complexity. The ISODATA algorithm adapted for our purpose works well when we cluster speakers whose acoustic characteristics are similar to a distance measure. The time spent on HSI (hierarchical speaker identification) is approximately 30.3 percent more than that spent on CSI (conventional speaker identification) when the number of registered speakers is 40 in our experiments. Increasing of the number of speakers decreases the time spent on HSI compared with CSI. It is shown that this approach can improve the speed of the speaker identification system.Keywords
This publication has 2 references indexed in Scilit:
- A distance measure between collections of distributions and its application to speaker recognition2002
- Robust text-independent speaker identification using Gaussian mixture speaker modelsIEEE Transactions on Speech and Audio Processing, 1995