Blind adaptive identification of a Single-Input Multiple-Output (SIMO) acoustic system has useful applications including acoustic environment sensing, source localization and, in combination with multichannel equalization, dereverberation. An empirically chosen step-size is usually employed in blind system identification algorithms based on cross-relation error minimization. Although some adaptive step-size approaches have been proposed in the literature, the derivations rely, in some cases, on coarse approximations. In this paper, a locally optimal adaptive-step size exploiting the algebraic nature of the problem is derived. Experimental results using simulated room impulse responses show that the proposed algorithm has higher initial convergence rate.

Authors: Hu, Mathieu; Sharma, Dushyant; Doclo, Simon; Brookes, Mike; Naylor, Patrick A.
Affiliations: Department of Medical Physics and Acoustics, University of Oldenburg, Oldenburg, Germany; Imperial College London, London, UK; Voicemail-To-Text Research, Nuance Communications Inc., Marlow, UK(See document for exact affiliation information.)
AES Conference: 60th International Conference: DREAMS (Dereverberation and Reverberation of Audio, Music, and Speech) (January 2016)
Paper Number: 6-2
Publication Date: January 27, 2016 Import into BibTeX
Subject: Paper Session 6
Permalink: https://www.aes.org/e-lib/browse.cfm?elib=18082

Click to purchase paper as a non-member or login as an AES member. If your company or school subscribes to the E-Library then switch to the institutional version. If you are not an AES member and would like to subscribe to the E-Library then Join the AES!

This paper costs $33 for non-members and is free for AES members and E-Library subscribers.

Learn more about the AES E-Library

E-Library Location: (CD 60thPapers)   /conf/60/aes60-000041.pdf

Start a discussion about this paper!