Single-channel speech dereverberation in noisy environment for non-orthogonal signals
-
Altmetric Citations
Fahim, Abdullah; Samarasinghe, Prasanga; Abhayapala, Thushara
Description
The detrimental effect of speech reverberation reduces speech quality, limits the performance of automatic speech recognition systems and impairs hearing aids. Spectral enhancement (SE) is a popular method for suppressing the late reverberation and background noise. However, conventional SE-based approaches assume orthogonality between the desired and undesired signal components. This orthogonality assumption does not hold true in most of the practical cases due to a limited time-domain support...[Show more]
dc.contributor.author | Fahim, Abdullah | |
---|---|---|
dc.contributor.author | Samarasinghe, Prasanga | |
dc.contributor.author | Abhayapala, Thushara | |
dc.date.accessioned | 2020-07-10T00:24:20Z | |
dc.date.available | 2020-07-10T00:24:20Z | |
dc.identifier.issn | 1610-1928 | |
dc.identifier.uri | http://hdl.handle.net/1885/205999 | |
dc.description.abstract | The detrimental effect of speech reverberation reduces speech quality, limits the performance of automatic speech recognition systems and impairs hearing aids. Spectral enhancement (SE) is a popular method for suppressing the late reverberation and background noise. However, conventional SE-based approaches assume orthogonality between the desired and undesired signal components. This orthogonality assumption does not hold true in most of the practical cases due to a limited time-domain support and the short-time stationarity of the speech signals, and thereby, affects estimation accuracy. To circumvent this issue, Lu et al. relaxed the orthogonality assumption by proposing a geometric approach to spectral subtraction (GSS) and evaluated their algorithm against different kinds of background noise. In our work, we comprehensively analyze the model by virtue of a simplified GSS transfer function to gain an insight into the algorithm. We conduct a series of experiments to validate GSS and explore its limitations in diverse realistic scenarios with both reverberation and background noise through a comprehensive end-to-end system for speech dereverberation and noise suppression. We also analyze the performance of GSS using the experimental data of the 2014 REVERB challenge and compare it with other conventional approaches such as spectral subtraction, Wiener Filter, minimum mean square error short-time spectral amplitude estimator and log spectral amplitude estimator, as well as with the contemporary methods of the 2014 REVERB challenge. | |
dc.description.sponsorship | Australian Research Council: DP140103412 | |
dc.format.mimetype | application/pdf | |
dc.language.iso | en_AU | |
dc.publisher | S. Hirzel Verlag | |
dc.rights | © 2018 S. Hirzel Verlag | |
dc.source | Acta Acustica united with Acustica | |
dc.title | Single-channel speech dereverberation in noisy environment for non-orthogonal signals | |
dc.type | Journal article | |
local.description.notes | Imported from ARIES | |
local.identifier.citationvolume | 104 | |
dc.date.issued | 2018-11-01 | |
local.identifier.absfor | 090609 - Signal Processing | |
local.identifier.ariespublication | u4485658xPUB2658 | |
local.publisher.url | https://www.ingentaconnect.com/ | |
local.type.status | Accepted Version | |
local.contributor.affiliation | Fahim, Abdullah, College of Engineering and Computer Science, ANU | |
local.contributor.affiliation | Samarasinghe, Prasanga, College of Engineering and Computer Science, ANU | |
local.contributor.affiliation | Abhayapala, Pallage (Thushara), College of Engineering and Computer Science, ANU | |
dc.relation | http://purl.org/au-research/grants/arc/DP140103412 | |
local.bibliographicCitation.issue | 6 | |
local.bibliographicCitation.startpage | 1041 | |
local.bibliographicCitation.lastpage | 1055 | |
local.identifier.doi | 10.3813/AAA.919270 | |
local.identifier.absseo | 970109 - Expanding Knowledge in Engineering | |
dc.date.updated | 2020-06-23T00:56:06Z | |
local.identifier.scopusID | 2-s2.0-85056764189 | |
dcterms.accessRights | Open Access | |
dc.provenance | http://v2.sherpa.ac.uk/id/publication/23848..."Author Accepted Manuscript can be made open access on insitutional repository after 12 month embargo" from SHERPA/RoMEO site (as at 10/7/20). | |
Collections | ANU Research Publications |
Download
File | Description | Size | Format | Image |
---|---|---|---|---|
01_Fahim_Single-channel_speech_2018.pdf | 1.09 MB | Adobe PDF |
Items in Open Research are protected by copyright, with all rights reserved, unless otherwise indicated.
Updated: 17 November 2022/ Responsible Officer: University Librarian/ Page Contact: Library Systems & Web Coordinator