Perceptual quality of audio separated using sigmoidal masks. Stokes, T., Hummersone, C., Brookes, T., & Mason, A. In Proceedings of the 137th Audio Engineering Society Convention, Los Angeles, October, 2014. Paper abstract bibtex Separation of underdetermined audio mixtures is often performed in the Time-Frequency (TF) domain by masking each TF element according to the amount of target energy it is deemed to contain. This work uses sigmoidal functions to map the proportion of target energy to mask values. The series of sigmoidal functions used encompasses the ratio mask and an approximation of the binary mask. Mixtures are chosen to represent a range of different amounts of TF overlap, then separated and evaluated using objective measures. PEASS results show improved interferer suppression and artifact scores can be achieved using softer masking than that applied by binary or ratio masks. This improves the overall perceptual score of the separated audio.
@inproceedings{Stokes:2014,
Abstract = {Separation of underdetermined audio mixtures is often performed in the
Time-Frequency (TF) domain by masking each TF element according to the
amount of target energy it is deemed to contain. This work uses sigmoidal
functions to map the proportion of target energy to mask values. The series
of sigmoidal functions used encompasses the ratio mask and an approximation
of the binary mask. Mixtures are chosen to represent a range of different
amounts of TF overlap, then separated and evaluated using objective
measures. PEASS results show improved interferer suppression and artifact
scores can be achieved using softer masking than that applied by binary or
ratio masks. This improves the overall perceptual score of the separated
audio.},
Author = {Toby Stokes and Chris Hummersone and Tim Brookes and Andrew Mason},
Booktitle = {Proceedings of the 137th Audio Engineering Society Convention, Los Angeles},
Date-Added = {2015-09-25 14:44:56 +0000},
Date-Modified = {2015-09-25 14:44:56 +0000},
Keywords = {source separation; binary mask; ratio mask; sigmoidal mask; Wiener filter},
Month = {October},
Title = {Perceptual quality of audio separated using sigmoidal masks},
Year = {2014},
url_Paper = {http://tobystokes.s3-eu-west-1.amazonaws.com/docs/17505.pdf}}
Downloads: 0
{"_id":"yRt6hAdcjcri39P2h","bibbaseid":"stokes-hummersone-brookes-mason-perceptualqualityofaudioseparatedusingsigmoidalmasks-2014","downloads":0,"creationDate":"2015-09-25T14:48:26.508Z","title":"Perceptual quality of audio separated using sigmoidal masks","author_short":["Stokes, T.","Hummersone, C.","Brookes, T.","Mason, A."],"year":2014,"bibtype":"inproceedings","biburl":"https://s3-eu-west-1.amazonaws.com/tobystokes/TobyStokesPublications.bib","bibdata":{"bibtype":"inproceedings","type":"inproceedings","abstract":"Separation of underdetermined audio mixtures is often performed in the Time-Frequency (TF) domain by masking each TF element according to the amount of target energy it is deemed to contain. This work uses sigmoidal functions to map the proportion of target energy to mask values. The series of sigmoidal functions used encompasses the ratio mask and an approximation of the binary mask. Mixtures are chosen to represent a range of different amounts of TF overlap, then separated and evaluated using objective measures. PEASS results show improved interferer suppression and artifact scores can be achieved using softer masking than that applied by binary or ratio masks. This improves the overall perceptual score of the separated audio.","author":[{"firstnames":["Toby"],"propositions":[],"lastnames":["Stokes"],"suffixes":[]},{"firstnames":["Chris"],"propositions":[],"lastnames":["Hummersone"],"suffixes":[]},{"firstnames":["Tim"],"propositions":[],"lastnames":["Brookes"],"suffixes":[]},{"firstnames":["Andrew"],"propositions":[],"lastnames":["Mason"],"suffixes":[]}],"booktitle":"Proceedings of the 137th Audio Engineering Society Convention, Los Angeles","date-added":"2015-09-25 14:44:56 +0000","date-modified":"2015-09-25 14:44:56 +0000","keywords":"source separation; binary mask; ratio mask; sigmoidal mask; Wiener filter","month":"October","title":"Perceptual quality of audio separated using sigmoidal masks","year":"2014","url_paper":"http://tobystokes.s3-eu-west-1.amazonaws.com/docs/17505.pdf","bibtex":"@inproceedings{Stokes:2014,\n\tAbstract = {Separation of underdetermined audio mixtures is often performed in the\nTime-Frequency (TF) domain by masking each TF element according to the\namount of target energy it is deemed to contain. This work uses sigmoidal\nfunctions to map the proportion of target energy to mask values. The series\nof sigmoidal functions used encompasses the ratio mask and an approximation\nof the binary mask. Mixtures are chosen to represent a range of different\namounts of TF overlap, then separated and evaluated using objective\nmeasures. PEASS results show improved interferer suppression and artifact\nscores can be achieved using softer masking than that applied by binary or\nratio masks. This improves the overall perceptual score of the separated\naudio.},\n\tAuthor = {Toby Stokes and Chris Hummersone and Tim Brookes and Andrew Mason},\n\tBooktitle = {Proceedings of the 137th Audio Engineering Society Convention, Los Angeles},\n\tDate-Added = {2015-09-25 14:44:56 +0000},\n\tDate-Modified = {2015-09-25 14:44:56 +0000},\n\tKeywords = {source separation; binary mask; ratio mask; sigmoidal mask; Wiener filter},\n\tMonth = {October},\n\tTitle = {Perceptual quality of audio separated using sigmoidal masks},\n\tYear = {2014},\n\turl_Paper = {http://tobystokes.s3-eu-west-1.amazonaws.com/docs/17505.pdf}}\n\n","author_short":["Stokes, T.","Hummersone, C.","Brookes, T.","Mason, A."],"key":"Stokes:2014","id":"Stokes:2014","bibbaseid":"stokes-hummersone-brookes-mason-perceptualqualityofaudioseparatedusingsigmoidalmasks-2014","role":"author","urls":{" paper":"http://tobystokes.s3-eu-west-1.amazonaws.com/docs/17505.pdf"},"keyword":["source separation; binary mask; ratio mask; sigmoidal mask; Wiener filter"],"downloads":0,"html":""},"search_terms":["perceptual","quality","audio","separated","using","sigmoidal","masks","stokes","hummersone","brookes","mason"],"keywords":["source separation; binary mask; ratio mask; sigmoidal mask; wiener filter"],"authorIDs":["56055eba80505273060001fa"],"dataSources":["wyzvCX55eHzm6BkH2"]}