Publications-Detail
Performance Analysis of Wavelet-based Voice Activity Detection
- Authors:
- Jeub, M. , Kolossa, D. , Fernandez Astudillo, R. , Orglmeister, R.
- Book Title:
- Proceedings of the International Conference on Acoustics, including the 35th German Annual Conference on Acoustics (DAGA)
- Organization:
- DEGA
- Publisher:
- Deutsche Gesellsch. f. Akustik
- Pages:
- p.p. 407-410
- Address:
- Rotterdam, The Netherlands
- Date:
- Mar. 2009
- ISBN:
- 978-3-98086-596-8
- Language:
- English
Abstract
The objective of this paper is to analyze the performance of wavelet-based voice activity detection algorithms (VAD) and to contrast it with that of the VAD standardized for the AMR-WB (Adaptive Multi-Rate Wideband) codec. Experimental results in clean, noisy and reverberant environments show that wavelet approaches lead to good results with respect to speech clipping and offer a much lower computational complexity. Integration of these algorithms into a Hidden Markov model (HMM) speech recognizer shows that the recognition performance using the AMR VAD can also be obtained or improved upon by wavelet based approaches, again at a notably reduced computational effort.Download
BibTeX- Copyright © by IKS
jeub09b.pdf - This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.