Abstract
This paper presents an automatic gender recognition technique based on speaker's voice. Utterances spoken by same-gender speakers were used to train a text-independent hidden Markov gender model. female and male models are continous hidden Markov models. Experiments on the TI-46 database containing 46 words spoken by 8 female and 8 male speakers showed an acceptable recognition rate.
Original language | English |
---|---|
Title of host publication | Proceedings of the 2nd WSEAS International Conference on Electronics, Control and Signal Processing (ICECS '03) |
Editors | Nikos Mastorakis |
Place of Publication | Singapore |
Publisher | WSEAS Transactions |
Pages | 1-5 |
Number of pages | 5 |
ISBN (Print) | 960-8052-91-2 |
Publication status | Published - 2003 |
Event | 2nd WSEAS International Conference - , Singapore Duration: 7 Dec 2003 → 9 Dec 2003 |
Conference
Conference | 2nd WSEAS International Conference |
---|---|
Country/Territory | Singapore |
Period | 7/12/03 → 9/12/03 |