Effectiveness of ideal ratio mask for non-intrusive quality assessment of noise suppressed speech. Soni, M. H. & Patil, H. A. In 2017 25th European Signal Processing Conference (EUSIPCO), pages 573-577, Aug, 2017. Paper doi abstract bibtex The Ideal Ratio Mask (IRM) has proven to be very effective tool in many applications such as speech segregation, speech enhancement for hearing aid design and noise robust speech recognition tasks. The IRM provides information regarding the amount of signal power at each Time-Frequency (T-F) unit in a given signal-plus-noise mixture. In this paper, we propose to use the IRM for non-intrusive quality assessment of noise suppressed speech. Since the quality of noise suppressed speech is dependent on the residual noise present in speech, IRM can be extremely useful for its quality assessment. The quality assessment problem is posed as a regression problem and the mapping between statistics of acoustic features, namely, Mel Filterbank Energies (FBEs) plus IRM features and the subjective score of the corresponding utterances was found using single-layer Artificial Neural Network (ANN). The results of our experiments suggest that by using the mean of FBEs and IRM features as the input, the quality prediction accuracy was significantly increased.
@InProceedings{8081272,
author = {M. H. Soni and H. A. Patil},
booktitle = {2017 25th European Signal Processing Conference (EUSIPCO)},
title = {Effectiveness of ideal ratio mask for non-intrusive quality assessment of noise suppressed speech},
year = {2017},
pages = {573-577},
abstract = {The Ideal Ratio Mask (IRM) has proven to be very effective tool in many applications such as speech segregation, speech enhancement for hearing aid design and noise robust speech recognition tasks. The IRM provides information regarding the amount of signal power at each Time-Frequency (T-F) unit in a given signal-plus-noise mixture. In this paper, we propose to use the IRM for non-intrusive quality assessment of noise suppressed speech. Since the quality of noise suppressed speech is dependent on the residual noise present in speech, IRM can be extremely useful for its quality assessment. The quality assessment problem is posed as a regression problem and the mapping between statistics of acoustic features, namely, Mel Filterbank Energies (FBEs) plus IRM features and the subjective score of the corresponding utterances was found using single-layer Artificial Neural Network (ANN). The results of our experiments suggest that by using the mean of FBEs and IRM features as the input, the quality prediction accuracy was significantly increased.},
keywords = {acoustic noise;acoustic signal processing;channel bank filters;feature extraction;neural nets;regression analysis;speech enhancement;time-frequency analysis;signal-plus-noise mixture;time-frequency unit;T-F unit;acoustic features;mel filterbank energies;FBE;artificial neural network;ANN;regression problem;nonintrusive quality assessment;ideal ratio mask;quality assessment problem;IRM;residual noise;noise suppressed speech;Speech;Feature extraction;Quality assessment;Noise measurement;Acoustics;Reactive power},
doi = {10.23919/EUSIPCO.2017.8081272},
issn = {2076-1465},
month = {Aug},
url = {https://www.eurasip.org/proceedings/eusipco/eusipco2017/papers/1570339953.pdf},
}
Downloads: 0
{"_id":"rMnATsdPspLzTLLWv","bibbaseid":"soni-patil-effectivenessofidealratiomaskfornonintrusivequalityassessmentofnoisesuppressedspeech-2017","authorIDs":[],"author_short":["Soni, M. H.","Patil, H. A."],"bibdata":{"bibtype":"inproceedings","type":"inproceedings","author":[{"firstnames":["M.","H."],"propositions":[],"lastnames":["Soni"],"suffixes":[]},{"firstnames":["H.","A."],"propositions":[],"lastnames":["Patil"],"suffixes":[]}],"booktitle":"2017 25th European Signal Processing Conference (EUSIPCO)","title":"Effectiveness of ideal ratio mask for non-intrusive quality assessment of noise suppressed speech","year":"2017","pages":"573-577","abstract":"The Ideal Ratio Mask (IRM) has proven to be very effective tool in many applications such as speech segregation, speech enhancement for hearing aid design and noise robust speech recognition tasks. The IRM provides information regarding the amount of signal power at each Time-Frequency (T-F) unit in a given signal-plus-noise mixture. In this paper, we propose to use the IRM for non-intrusive quality assessment of noise suppressed speech. Since the quality of noise suppressed speech is dependent on the residual noise present in speech, IRM can be extremely useful for its quality assessment. The quality assessment problem is posed as a regression problem and the mapping between statistics of acoustic features, namely, Mel Filterbank Energies (FBEs) plus IRM features and the subjective score of the corresponding utterances was found using single-layer Artificial Neural Network (ANN). The results of our experiments suggest that by using the mean of FBEs and IRM features as the input, the quality prediction accuracy was significantly increased.","keywords":"acoustic noise;acoustic signal processing;channel bank filters;feature extraction;neural nets;regression analysis;speech enhancement;time-frequency analysis;signal-plus-noise mixture;time-frequency unit;T-F unit;acoustic features;mel filterbank energies;FBE;artificial neural network;ANN;regression problem;nonintrusive quality assessment;ideal ratio mask;quality assessment problem;IRM;residual noise;noise suppressed speech;Speech;Feature extraction;Quality assessment;Noise measurement;Acoustics;Reactive power","doi":"10.23919/EUSIPCO.2017.8081272","issn":"2076-1465","month":"Aug","url":"https://www.eurasip.org/proceedings/eusipco/eusipco2017/papers/1570339953.pdf","bibtex":"@InProceedings{8081272,\n author = {M. H. Soni and H. A. Patil},\n booktitle = {2017 25th European Signal Processing Conference (EUSIPCO)},\n title = {Effectiveness of ideal ratio mask for non-intrusive quality assessment of noise suppressed speech},\n year = {2017},\n pages = {573-577},\n abstract = {The Ideal Ratio Mask (IRM) has proven to be very effective tool in many applications such as speech segregation, speech enhancement for hearing aid design and noise robust speech recognition tasks. The IRM provides information regarding the amount of signal power at each Time-Frequency (T-F) unit in a given signal-plus-noise mixture. In this paper, we propose to use the IRM for non-intrusive quality assessment of noise suppressed speech. Since the quality of noise suppressed speech is dependent on the residual noise present in speech, IRM can be extremely useful for its quality assessment. The quality assessment problem is posed as a regression problem and the mapping between statistics of acoustic features, namely, Mel Filterbank Energies (FBEs) plus IRM features and the subjective score of the corresponding utterances was found using single-layer Artificial Neural Network (ANN). The results of our experiments suggest that by using the mean of FBEs and IRM features as the input, the quality prediction accuracy was significantly increased.},\n keywords = {acoustic noise;acoustic signal processing;channel bank filters;feature extraction;neural nets;regression analysis;speech enhancement;time-frequency analysis;signal-plus-noise mixture;time-frequency unit;T-F unit;acoustic features;mel filterbank energies;FBE;artificial neural network;ANN;regression problem;nonintrusive quality assessment;ideal ratio mask;quality assessment problem;IRM;residual noise;noise suppressed speech;Speech;Feature extraction;Quality assessment;Noise measurement;Acoustics;Reactive power},\n doi = {10.23919/EUSIPCO.2017.8081272},\n issn = {2076-1465},\n month = {Aug},\n url = {https://www.eurasip.org/proceedings/eusipco/eusipco2017/papers/1570339953.pdf},\n}\n\n","author_short":["Soni, M. H.","Patil, H. A."],"key":"8081272","id":"8081272","bibbaseid":"soni-patil-effectivenessofidealratiomaskfornonintrusivequalityassessmentofnoisesuppressedspeech-2017","role":"author","urls":{"Paper":"https://www.eurasip.org/proceedings/eusipco/eusipco2017/papers/1570339953.pdf"},"keyword":["acoustic noise;acoustic signal processing;channel bank filters;feature extraction;neural nets;regression analysis;speech enhancement;time-frequency analysis;signal-plus-noise mixture;time-frequency unit;T-F unit;acoustic features;mel filterbank energies;FBE;artificial neural network;ANN;regression problem;nonintrusive quality assessment;ideal ratio mask;quality assessment problem;IRM;residual noise;noise suppressed speech;Speech;Feature extraction;Quality assessment;Noise measurement;Acoustics;Reactive power"],"metadata":{"authorlinks":{}},"downloads":0},"bibtype":"inproceedings","biburl":"https://raw.githubusercontent.com/Roznn/EUSIPCO/main/eusipco2017url.bib","creationDate":"2021-02-13T16:38:25.554Z","downloads":0,"keywords":["acoustic noise;acoustic signal processing;channel bank filters;feature extraction;neural nets;regression analysis;speech enhancement;time-frequency analysis;signal-plus-noise mixture;time-frequency unit;t-f unit;acoustic features;mel filterbank energies;fbe;artificial neural network;ann;regression problem;nonintrusive quality assessment;ideal ratio mask;quality assessment problem;irm;residual noise;noise suppressed speech;speech;feature extraction;quality assessment;noise measurement;acoustics;reactive power"],"search_terms":["effectiveness","ideal","ratio","mask","non","intrusive","quality","assessment","noise","suppressed","speech","soni","patil"],"title":"Effectiveness of ideal ratio mask for non-intrusive quality assessment of noise suppressed speech","year":2017,"dataSources":["2MNbFYjMYTD6z7ExY","uP2aT6Qs8sfZJ6s8b"]}