Skip navigation
Skip navigation

Spatial feature learning for robust binaural sound source localization using a composite feature vector

Wu, Xiang; Talagala, Dumidu S.; Zhang, Wen; Abhayapala, Thushara

Description

The performance of binaural speech source localization systems can be significantly impacted by an imperfect selection of spatial localization cues, due to the limited bandwidth of speech, and the effects of noise. In order to mitigate these impacts, this paper presents a novel method that combines a deterministic localization approach with a spatial feature learning process. Here, we (i) obtain a composite feature vector derived from analysing the mutual information between different spatial...[Show more]

dc.contributor.authorWu, Xiang
dc.contributor.authorTalagala, Dumidu S.
dc.contributor.authorZhang, Wen
dc.contributor.authorAbhayapala, Thushara
dc.date.accessioned2016-09-01T03:56:52Z
dc.date.available2016-09-01T03:56:52Z
dc.identifier.isbn978-1-4799-9988-0
dc.identifier.issn1520-6149
dc.identifier.urihttp://hdl.handle.net/1885/107954
dc.description.abstractThe performance of binaural speech source localization systems can be significantly impacted by an imperfect selection of spatial localization cues, due to the limited bandwidth of speech, and the effects of noise. In order to mitigate these impacts, this paper presents a novel method that combines a deterministic localization approach with a spatial feature learning process. Here, we (i) obtain a composite feature vector derived from analysing the mutual information between different spatial cues and (ii) estimate the optimum feature combination that minimizes the angular localization error in three dimensional space. The performance of the proposed mutual information based feature learning approach is evaluated for a range of speech inputs and noise conditions. We also demonstrate that the proposed approach improves the localization accuracy and its robustness, with respect to traditional localization algorithms, especially in the relatively low signal-to-noise ratio localization scenarios.
dc.description.sponsorshipThis work was supported under the Australian Research Councils Discovery Projects funding scheme (project no. DE150100363).
dc.publisherInstitute of Electrical and Electronics Engineers (IEEE)
dc.relation.ispartof2016 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 20-25 March 2016
dc.rights© 2016 IEEE
dc.titleSpatial feature learning for robust binaural sound source localization using a composite feature vector
dc.typeConference paper
dc.date.issued2016
local.publisher.urlhttp://www.ieee.org/index.html
local.type.statusPublished Version
local.contributor.affiliationWu, X., Research School of Engineering, CECS, The Australian National University
local.contributor.affiliationZhang, W., Research School of Engineering, CECS, The Australian National University
local.contributor.affiliationAbhayapala, T. D., Research School of Engineering, CECS, The Australian National University
dc.relationhttp://purl.org/au-research/grants/arc/DE150100363
local.identifier.essn2379-190X
local.bibliographicCitation.startpage6320
local.bibliographicCitation.lastpage6324
local.identifier.doi10.1109/ICASSP.2016.7472893
CollectionsANU Research Publications

Download

There are no files associated with this item.


Items in Open Research are protected by copyright, with all rights reserved, unless otherwise indicated.

Updated:  19 May 2020/ Responsible Officer:  University Librarian/ Page Contact:  Library Systems & Web Coordinator