A kepstrum approach to filtering, smoothing and prediction with application to speech enhancement

Abstract

Restricted accessMoreSectionsView PDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareShare onFacebookTwitterLinked InRedditEmail Cite this article Moir T. J. and Barrett J. F. 2003A kepstrum approach to filtering, smoothing and prediction with application to speech enhancementProc. R. Soc. Lond. A.4592957–2976http://doi.org/10.1098/rspa.2003.1137SectionRestricted accessA kepstrum approach to filtering, smoothing and prediction with application to speech enhancement T. J. Moir T. J. Moir Institute of Information and Mathematical Sciences, Massey University at Albany, Private Bag 102–904, Auckland, New Zealand () Google Scholar Find this author on PubMed Search for more papers by this author and J. F. Barrett J. F. Barrett 40 Botley Road, Southampton SO19 0NS, UK Google Scholar Find this author on PubMed Search for more papers by this author T. J. Moir T. J. Moir Institute of Information and Mathematical Sciences, Massey University at Albany, Private Bag 102–904, Auckland, New Zealand () Google Scholar Find this author on PubMed Search for more papers by this author and J. F. Barrett J. F. Barrett 40 Botley Road, Southampton SO19 0NS, UK Google Scholar Find this author on PubMed Search for more papers by this author Published:08 December 2003https://doi.org/10.1098/rspa.2003.1137AbstractA kepstrum (or complex–cepstrum) approach to minimum–phase Wiener filtering of stationary scalar processes is proposed and solved for the case of signal plus coloured noise, where the noise possibly includes a white–noise component. A general solution is found in an innovations form. The spectral factorization of the noise model and of the signal–plus–noise model required for the solution are determined from data using the kepstrum technique with the fast Fourier transform. This approach avoids dependence on any form of multidimensional state–space or polynomial–based model and so avoids use of recursive parameter estimation or of Diophantine equations. Previous ArticleNext Article VIEW FULL TEXT DOWNLOAD PDF FiguresRelatedReferencesDetailsCited by Pan J, Xing A, Zhu J, Nong S, Ma Y, Zhu X, Fang W and Wang Y (2021) Gene expression analysis in leaf of Camellia sinensis reveals the response to fluoride, Acta Physiologiae Plantarum, 10.1007/s11738-021-03283-5, 43:7, Online publication date: 1-Jul-2021. Alar H, Mamaril R, Villegas L and Cabarrubias J (2021) Audio classification of violin bowing techniques: An aid for beginners, Machine Learning with Applications, 10.1016/j.mlwa.2021.100028, 4, (100028), Online publication date: 1-Jun-2021. Sharma G, Umapathy K and Krishnan S (2020) Trends in audio signal feature extraction methods, Applied Acoustics, 10.1016/j.apacoust.2019.107020, 158, (107020), Online publication date: 1-Jan-2020. Nason G, Powell B, Elliott D and Smith P (2016) Should we sample a time series more frequently?: decision support via multirate spectrum estimation, Journal of the Royal Statistical Society: Series A (Statistics in Society), 10.1111/rssa.12210, 180:2, (353-407), Online publication date: 1-Feb-2017. Moir T (2014) Spectral factorization using FFTs for large-scale problems, International Journal of Adaptive Control and Signal Processing, 10.1002/acs.2512, 29:8, (954-970), Online publication date: 1-Aug-2015. Moir T (2012) Filtering, smoothing and prediction using a control-loop spectral factorization method for coloured noise, International Journal of Adaptive Control and Signal Processing, 10.1002/acs.2285, 27:3, (153-165), Online publication date: 1-Mar-2013. Lalitha V, Prema P and Mathew L (2010) A Kepstrum based approach for enhancement of dysarthric speech 2010 3rd International Congress on Image and Signal Processing (CISP), 10.1109/CISP.2010.5646752, 978-1-4244-6513-2, (3474-3478) Jeong J (2010) Real-time acoustic noise canceling technique on innovations-based inverse kepstrum and FIR RLS Control (MSC), 10.1109/ISIC.2010.5612882, 978-1-4244-5360-3, (2444-2449) Jeong J (2009) Analysis of inverse kepstrum and innovations-based application to noise cancellation 2009 IEEE International Symposium on Industrial Electronics (ISIE 2009), 10.1109/ISIE.2009.5217921, 978-1-4244-4347-5, (890-896) Moir T (2008) A kepstrum approach to time-delay estimation with application to acoustic source tracking 2008 15th International Conference on Mechatronics and Machine Vision in Practice (M2VIP), 10.1109/MMVIP.2008.4749563, 978-1-4244-3779-5, (378-384) Jeong J and Moir T (2008) A real-time kepstrum approach to speech enhancement and noise cancellation, Neurocomputing, 10.1016/j.neucom.2007.09.026, 71:13-15, (2635-2649), Online publication date: 1-Aug-2008. Moir T (2007) A z -domain transfer function solution to the non-minimum phase acoustic beamformer , International Journal of Systems Science, 10.1080/00207720701431276, 38:7, (563-575), Online publication date: 1-Jul-2007. Moir T and Jeong J (2006) Identification of Non-Minimum Phase Transfer Function Components 2006 IEEE International Symposium on Signal Processing and Information Technology, 10.1109/ISSPIT.2006.270830, 0-7803-9754-1, (380-384) Jeong J (2011) An Innovations-Based Noise Cancelling Technique on Inverse Kepstrum Whitening Filter and Adaptive FIR Filter in Beamforming Structure, Sensors, 10.3390/s110706816, 11:7, (6816-6841) This Issue08 December 2003Volume 459Issue 2040 Article InformationDOI:https://doi.org/10.1098/rspa.2003.1137Published by:Royal SocietyPrint ISSN:1364-5021Online ISSN:1471-2946History: Published online08/12/2003Published in print08/12/2003 License: Citations and impact KeywordsColoured NoiseSmoothingComplex CepstrumPredictionKepstrum