| Paper: | SLP-P12.3 | ||
| Session: | Speech Processing for Reverberation, Quantization and Enhancement | ||
| Time: | Thursday, May 18, 10:00 - 12:00 | ||
| Presentation: | Poster | ||
| Topic: | Speech and Spoken Language Processing: Speech Enhancement (for Impaired Situations) | ||
| Title: | Speech dereverberation based on probabilistic models of source and room acoustics | ||
| Authors: | Tomohiro Nakatani, NTT Corporation, Japan; Biing-Hwang Juang, Georgia Institute of Technology, United States; Keisuke Kinoshita, Masato Miyoshi, NTT Corporation, Japan | ||
| Abstract: | This paper proposes a new single channel speech dereverberation method, in which the features of speech and room acoustics are modeled by probabilistic density functions (pdf), and the source signals are estimated by maximizing a likelihood function defined based on the pdfs. Two types of pdfs are introduced for the source signals based on harmonicity and sparseness, while the pdf for the room acoustics is defined based on an inverse filtering operation. The EM algorithm is used to solve this maximum likelihood problem efficiently. The effectiveness of the present method is shown in terms of the energy decay curves of the dereverberated impulse responses. | ||
©2018 Conference Management Services, Inc. -||- email: webmaster@icassp2006.org -||- Last updated Friday, August 17, 2012