Mixture IMM for speech enhancement under nonstationary noise

Authors
Kang, SK Baek, SJ Lee, KY Sung, KM
Citation
Sk. Kang et al., Mixture IMM for speech enhancement under nonstationary noise, IEEE SPEECH, 8(5), 2000, pp. 637-641
Citations number
6
Categorie Soggetti
Eletrical & Eletronics Engineeing
Journal title
IEEE TRANSACTIONS ON SPEECH AND AUDIO PROCESSING
ISSN journal
10636676 → ACNP
Volume
8
Issue
5
Year of publication
2000
Pages
637 - 641
Database
ISI
SICI code
1063-6676(200009)8:5<637:MIFSEU>2.0.ZU;2-Z
Abstract
A mixture interacting multiple model (MIMM) algorithm is proposed to enhanc e speech contaminated by additive nonstationary noise. In this approach, a mixture hidden filter model (HFM) is used for clean speech modeling and a s ingle hidden filter is used for noise process modeling. The MIMM algorithm gives better enhancement results than the IMM algorithm. The results show t hat the proposed method offers performance gain compared to the previous re sults in [2] with slightly increased complexity.