单通道盲信号分离研究进展与展望
|
摘要
盲信号分离是当前国际信号处理领域的研究热点,作为一种病态混叠情况,单通道盲信号分离是其刚刚兴起的一个重要研究方向,有着广泛的应用前景。首先阐述了单通道盲信号分离的数学模型、可分离性分析和分离效果评价指标,然后在深入研究单通道盲信号分离最新发展的基础上,对现有理论和方法进行了总结,并介绍了一些领域中的应用情况,最后指出其发展面临的挑战和进一步的研究方向。
BSS(Blind Signal Separation) is currently a hotspot in the domain of signal processing in international community.As a case of ill-conditioned signal mixing,Single Channel Blind Signal Separation(SCBSS) is an aspect of particular interest and importance,as well as a new area of research,with a great future lying ahead.The mathematical model used for SCBSS,its separability and separation effect evaluation index are described in this paper.On the basis of an in-depth investigation into latest development of SCBSS,existing theories and methods are then classified and summarized,their application in some fields systematically explained.Finally,challenges and prospects for further development in the area are presented.
BSS(Blind Signal Separation) is currently a hotspot in the domain of signal processing in international community.As a case of ill-conditioned signal mixing,Single Channel Blind Signal Separation(SCBSS) is an aspect of particular interest and importance,as well as a new area of research,with a great future lying ahead.The mathematical model used for SCBSS,its separability and separation effect evaluation index are described in this paper.On the basis of an in-depth investigation into latest development of SCBSS,existing theories and methods are then classified and summarized,their application in some fields systematically explained.Finally,challenges and prospects for further development in the area are presented.
引文
[1]张贤达,保铮.盲信号分离[J].电子学报,2001,29(12A):1 766-1 771.
[2]刘琚,何振亚.盲源分离与盲反卷积[J].电子学报,2002,30(4):570-576.
[3]李小军,朱孝龙,张贤达.盲信号分离研究分类与展望[J].西安电子科技大学学报(自然科学版),2004,31(3):399-404.
[4]丁志中.盲信号分离及其在心电和语音信号处理中的应用[D].合肥:中国科学技术大学,2007.
[5]黄青华.基于源信号模型的盲分离技术研究及应用[D].上海:上海交通大学,2007.
[6]高鹰,姚振坚,李朝晖,等.基于广义特征值和核函数的非线性盲分离算法[J].系统工程与电子技术,2006,28(10):1 490-1 493.
[7]张贤达.现代信号处理[M].第二版.北京:清华大学出版社,2002.
[8]XI-REN CAO,RUEY-WEN LIU.General Approach to BlindSource Separation[J].IEEE Transactions on Signal Pro-cessing,1996,44(3):562-571.
[9]BYOUNGSEON CHOI.Separability and Axial Symmetryof M-D Polynomial and Maximum-entropy M-D RandomProcesses[J].IEEE Signal Processing Letters,2004,11(2):193-196.
[10]LEON COHEN.What is a Multicomponent Signal?[C].ICASSP 1992,San Francisco,CA,USA,1992,5:113-116.
[11]JAMES R.HOPGOOD,PETER J W RAYNER.Single Chan-nel Nonstationary Stochastic Signal Separation Using LinearTime-varying Filters[J].IEEE Transactions on Signal Pro-cessing,2003,51(7):1 739-1 752.
[12]LARS KAI HANSEN,KAARE BRANDT PETERSEN.Mon-aural ICA of White Noise is Hard[C].ICA 2003,Nara,Ja-pan,2003,815-820.
[13]蔡权伟,魏平,肖先赐.信道重叠信号分离方法的发展与展望[J].电子学报,2005,33(12A):2 446-2 454.
[14]徐明彪,朱维彰.关于信号盲分离分离效果评判指标的分析[J].杭州电子工业学院学报,2002,22(3):63-66.
[15]GIL-JIN JANG,TE-WON LEE,YUNG-HWAN OH.SingleChannel Signal Separation Using MAP-based SubspaceDecomposition[J].Electronics Letters,2003,39(24):1 766-1 767.
[16]邓智恒,罗伟栋.实际语音盲分离客观评价指标研究[J].电声技术,2007,31(9):61-65.
[17]李合生,赵明生,吴佑寿.一种短波通信盲信道快速分离算法[J].通信学报,2003,24(7):59-65.
[18]王立世,张永清,莫金垣.连续样条小波变换解析高噪声重叠分析信号[J].华南理工大学学报(自然科学版),2005,33(2):79-81.
[19]刘云,郭洁,叶芝慧,等.频谱重叠信号分离的循环平稳算法[J].东南大学学报(自然科学版),2005,35(3):333-337.
[20]黄健冲,姜秋喜,祁建清.数字接收机对同时到达信号的软件分离技术[J].航天电子对抗,2001(4):42-44.
[21]朱贻盛,李丹,等.混叠周期信号的自适应分离[J].信号处理,1992,8(4):202-210.
[22]DANIELS BENINCASA.Simultaneous Adaptive Co-chan-nel Speaker Separation,AFRL-IF-RS-TR-1999-162[R].Rome,NewYork:Air Force Research Laboratory,Informa-tion Directorate,Public Affairs office(IFOIPA),1999.
[23]LES ATLAS,CHRISTIAAN JANSSEN.Coherent Modula-tion Spectral Filtering for Single-channel Music SourceSeparation[C].ICASSP 2005,Philadelphia,PA,USA,2005,4:461-464.
[24]ALEXEY OZEROV,PIERRICK PHILIPPE,R啨I GRIBON-VAL,et al.One Microphone Singing Voice SeparationUsing Source-adapted Models[C].2005 IEEE Workshopon Applications of Signal Processing to Audio and Acous-tics,NewPaltz,NY,USA,2005,90-93.
[25]CHRISTOPHER J JAMES,D LOWE.Single Channel A-nalysis of Electromagnetic Brain Signals Through ICA in aDynamical Systems Framework[C].EMBS 2001,Istan-bul,Turkey,2001,1 974-1 977.
[26]PING CAO,EE-CHIEN CHANG,LONCE WYSE.BlindSeparation of Fetal ECG from Single Mixture Using SVDand ICA[C].ICICS-PCM 2003,Singapore,2003,3:1 418-1 422.
[27]WARNER E S,PROUDLER I K.Single-channel Blind Sig-nal Separation of Filtered MPSK Signals[J].IEE Proceed-ings of Radar,Sonar&Navigation,2003,150(6):396-402.
[28]LEE J,LEE K J,YOO S K.Development of a New SignalProcessing Algorithm Based on Independent ComponentAnalysis for Single Channel ECG data[C].EMBS 2004,San Francisco,CA,USA,1:224-226.
[29]GIL-JIN JANG,TE-WON LEE,YUNG-HWAN OH.Sin-gle-channel Signal Separation Using Time-domain BasisFunctions[J].IEEE Signal Processing Letters,2003,10(6):168-171.
[30]SAMT.ROWEIS.One Microphone Source Separation[C].NIPS 2000,Denver,CO,USA,December,2000,793-799.
[31]THOMAS BEIERHOLM,BRIAN DAM PEDERSEN,OLEWINTHER.Low Complexity Bayesian Single ChannelSource Separation[C].ICASSP 2004,Montreal,Quebec,Canada,2004,5:529-532.
[32]MANUEL J.REYES-GOMEZ,DANIEL P W ELLIS,etal.Multiband Audio Modeling for Single-channel AcousticSource Separation[C].ICASSP 2004,Montreal,Quebec,Canada,2004,5:641-644.
[33]BHIKSHA RAJ,MADHUSUDANA V S SHASHANKA,PARIS SMARAGDIS.Latent Dirichlet Decomposition forSingle Channel Speaker Separation[C].ICASSP 2006,Toulouse,France,2006,5:821-824.
[34]TUOMAS VIRTANEN.Monaural Sound Source Separationby Nonnegative Matrix Factorization with Temporal Conti-nuity and Sparseness Criteria[J].IEEE Transactions onAudio,Speech,and Language Processing,2007,15(3):1 066-1 074.
[35]C啨DRIC F啨VOTTE,SIMON J GODSILL.A Bayesian Ap-proach for Blind Separation of Sparse Sources[J].IEEETransactions on Audio,Speech,and Language Processing,2006,14(6):2 174-2 188.
[36]徐尚志,苏勇,叶中付.欠定条件下的盲分离算法[J].数据采集与处理,2006,21(2):128-132.
[37]HAROLD SZU,PORNCHAI CHANYAGORN,IVICA KO-PRIVA.Sparse Coding Blind Source Separation ThroughPowerline[J].Neurocomputing,2002,48(1):1 015-1 020.
[38]邱伟,黄秀轩.计算听觉场景分析介绍[J].高技术通讯,2002,12(1):106-110.
[39]DELIANG WANG,GUY J BROWN.Separation of Speechfrom Interfering Sounds Based on Oscillatory Correlation[J].IEEE Transactions on Neural Networks,1999,10(3):684-697.
[40]GUY J BROWN,DELIANG WANG.Separation of Speechby Computational Auditory Scene Analysis[C]//J Benes-ty,S Makino,J Chen.Speech Enhancement:What’s New?New York,USA:Springer,2005,371-402.
[41]MINGYANG WU,DELIANG WANG,GUY J BROWN.AMultipitch Tracking Algorithm for Noisy Speech[J].IEEETransactions on Speech and Audio Processing,2003,11(3):229-241.
[42]GUONING HU,DELIANG WANG.Monaural Speech Seg-regation Based on Pitch Tracking and Amplitude Modula-tion[J].IEEE Transactions on Neural Networks,2004,15(5):1 135-1 150.
[43]PENG LI,YONG GUAN,BO XU,et al.Monaural SpeechSeparation Based on Computational Auditory Scene Analy-sis and Objective Quality Assessment of Speech[J].IEEETransactions on Audio,Speech,and Language Processing,2006,14(6):2 014-2 023.
[44]MOHAMMAD HRADFAR,RICHARD MDANSEREAU,ABOLGHASEM SAYADIYAN.A Maximum LikelihoodEstimation of Vocal-tract-related Filter Characteristics forSingle Channel Speech Separation[J].EURASIP Journalon Audio,Speech,and Music Processing,2007(1):1-15.
[45]张晓冬,王桥,吴乐南.利用脊的特征进行信号盲分离[J].电子学报,2004,32(7):1 156-1 159.
[46]蔡权伟,魏平,肖先赐.基于模型拟合的重叠信号盲分离方法[J].电子学报,2005,33(10):1 794-1 798.
[47]EMMANUEL VINCENT,R GRIBONVAL,C FEVOTTE.Performance Measurement in Blind Audio Source Sepa-rateion[J].IEEE Transactions on Audio,Speech,andLanguage Processing,2006,14(4),1 462-1 469.
[48]蔡权伟,魏平,肖先赐.单信道多信号分量分离[J].通信学报,2006,27(6):49-56.
[49]CAI QUANWEI,WEI PING,XIAO XIANCI.Single Chan-nel Signal Component Separation Using Bayesian Estima-tion[J].Journal of Systems Engineering and Electronics,2007,18(1):33-39.
[50]张秀琦,刘辉,郑建斌,等.信号处理技术在重叠化学信号分析中的应用[J].化学进展,2002,14(3):174-181.
[51]ASHWIN RAO,RAMDAS KUMARESAN.Separation ofCochannel Signals Using the Parametric Constant ModulusAlgorithm[C].ICASSP 1996,Atlanta,Georgia,USA,1996,5:2 686-2 689.
[52]ARVIND V KEERTHI,JOHN J SHYNK.Separation of Co-channel Signals in TDMA Mobile Radio[J].IEEE Trans-actions on Signal Processing,1998,46(10):2 684-2 697.
[53]王正明,段晓君.信噪分离的频域方法[J].中国科学(E辑),1999,29(6):525-531.
[54]易东云,朱炬波,王正明.三频带信号分离的样条函数方法[J].电子学报,1999,27(1):54-57.
[55]周成,刘钊.基于独立分量分析的单通道语音降噪算法研究[J].激光与红外,2005,35(5):374-377.
[56]MASAKI KAWAKASTU.Application of ICAto MEG NoiseReduction[C].ICA2003,Nara,Japan,2003,535-541.
[57]OMAR A YESTE-OJEDA,JESUS GRAJAL.Cyclostation-arity-based Signal Separation in Interceptors Based on aSingle Sensor[C].2008 IEEE Radar Conference,RomeItaly,2008,169-174.
[58]SAM HEIDARI,CHRYSOSTOMOS L NIKIAS.Co-chan-nel Interference Mitigation in the Time-scale Domain:theCIMTS Algorithm[J].IEEE Transactions on Signal Pro-cessing,1996,44(9):2 151-2 162.
[59]CECILIA DI RUBERTO,ANDREW DEMPSTER,SHA-HID KHAN,et al.Segmentation of Blood Images UsingMorphological Operators[C].ICPR 2000,Barcelona,Spain,2000,3:397-400.
[60]郑宜珍,秦前清.小波包变换在重叠地震信号分离中的应用[J].计算机工程与应用,2004,40(26):176-177.
[61]HOONBIN HONG,MING LIANG.Separation of FaultFeatures from a Single-channel Mechanical Signal MixtureUsing Wavelet Decomposition[J].Mechanical Systems and Signal Processing,2007,21:2 025-2 040.
[2]刘琚,何振亚.盲源分离与盲反卷积[J].电子学报,2002,30(4):570-576.
[3]李小军,朱孝龙,张贤达.盲信号分离研究分类与展望[J].西安电子科技大学学报(自然科学版),2004,31(3):399-404.
[4]丁志中.盲信号分离及其在心电和语音信号处理中的应用[D].合肥:中国科学技术大学,2007.
[5]黄青华.基于源信号模型的盲分离技术研究及应用[D].上海:上海交通大学,2007.
[6]高鹰,姚振坚,李朝晖,等.基于广义特征值和核函数的非线性盲分离算法[J].系统工程与电子技术,2006,28(10):1 490-1 493.
[7]张贤达.现代信号处理[M].第二版.北京:清华大学出版社,2002.
[8]XI-REN CAO,RUEY-WEN LIU.General Approach to BlindSource Separation[J].IEEE Transactions on Signal Pro-cessing,1996,44(3):562-571.
[9]BYOUNGSEON CHOI.Separability and Axial Symmetryof M-D Polynomial and Maximum-entropy M-D RandomProcesses[J].IEEE Signal Processing Letters,2004,11(2):193-196.
[10]LEON COHEN.What is a Multicomponent Signal?[C].ICASSP 1992,San Francisco,CA,USA,1992,5:113-116.
[11]JAMES R.HOPGOOD,PETER J W RAYNER.Single Chan-nel Nonstationary Stochastic Signal Separation Using LinearTime-varying Filters[J].IEEE Transactions on Signal Pro-cessing,2003,51(7):1 739-1 752.
[12]LARS KAI HANSEN,KAARE BRANDT PETERSEN.Mon-aural ICA of White Noise is Hard[C].ICA 2003,Nara,Ja-pan,2003,815-820.
[13]蔡权伟,魏平,肖先赐.信道重叠信号分离方法的发展与展望[J].电子学报,2005,33(12A):2 446-2 454.
[14]徐明彪,朱维彰.关于信号盲分离分离效果评判指标的分析[J].杭州电子工业学院学报,2002,22(3):63-66.
[15]GIL-JIN JANG,TE-WON LEE,YUNG-HWAN OH.SingleChannel Signal Separation Using MAP-based SubspaceDecomposition[J].Electronics Letters,2003,39(24):1 766-1 767.
[16]邓智恒,罗伟栋.实际语音盲分离客观评价指标研究[J].电声技术,2007,31(9):61-65.
[17]李合生,赵明生,吴佑寿.一种短波通信盲信道快速分离算法[J].通信学报,2003,24(7):59-65.
[18]王立世,张永清,莫金垣.连续样条小波变换解析高噪声重叠分析信号[J].华南理工大学学报(自然科学版),2005,33(2):79-81.
[19]刘云,郭洁,叶芝慧,等.频谱重叠信号分离的循环平稳算法[J].东南大学学报(自然科学版),2005,35(3):333-337.
[20]黄健冲,姜秋喜,祁建清.数字接收机对同时到达信号的软件分离技术[J].航天电子对抗,2001(4):42-44.
[21]朱贻盛,李丹,等.混叠周期信号的自适应分离[J].信号处理,1992,8(4):202-210.
[22]DANIELS BENINCASA.Simultaneous Adaptive Co-chan-nel Speaker Separation,AFRL-IF-RS-TR-1999-162[R].Rome,NewYork:Air Force Research Laboratory,Informa-tion Directorate,Public Affairs office(IFOIPA),1999.
[23]LES ATLAS,CHRISTIAAN JANSSEN.Coherent Modula-tion Spectral Filtering for Single-channel Music SourceSeparation[C].ICASSP 2005,Philadelphia,PA,USA,2005,4:461-464.
[24]ALEXEY OZEROV,PIERRICK PHILIPPE,R啨I GRIBON-VAL,et al.One Microphone Singing Voice SeparationUsing Source-adapted Models[C].2005 IEEE Workshopon Applications of Signal Processing to Audio and Acous-tics,NewPaltz,NY,USA,2005,90-93.
[25]CHRISTOPHER J JAMES,D LOWE.Single Channel A-nalysis of Electromagnetic Brain Signals Through ICA in aDynamical Systems Framework[C].EMBS 2001,Istan-bul,Turkey,2001,1 974-1 977.
[26]PING CAO,EE-CHIEN CHANG,LONCE WYSE.BlindSeparation of Fetal ECG from Single Mixture Using SVDand ICA[C].ICICS-PCM 2003,Singapore,2003,3:1 418-1 422.
[27]WARNER E S,PROUDLER I K.Single-channel Blind Sig-nal Separation of Filtered MPSK Signals[J].IEE Proceed-ings of Radar,Sonar&Navigation,2003,150(6):396-402.
[28]LEE J,LEE K J,YOO S K.Development of a New SignalProcessing Algorithm Based on Independent ComponentAnalysis for Single Channel ECG data[C].EMBS 2004,San Francisco,CA,USA,1:224-226.
[29]GIL-JIN JANG,TE-WON LEE,YUNG-HWAN OH.Sin-gle-channel Signal Separation Using Time-domain BasisFunctions[J].IEEE Signal Processing Letters,2003,10(6):168-171.
[30]SAMT.ROWEIS.One Microphone Source Separation[C].NIPS 2000,Denver,CO,USA,December,2000,793-799.
[31]THOMAS BEIERHOLM,BRIAN DAM PEDERSEN,OLEWINTHER.Low Complexity Bayesian Single ChannelSource Separation[C].ICASSP 2004,Montreal,Quebec,Canada,2004,5:529-532.
[32]MANUEL J.REYES-GOMEZ,DANIEL P W ELLIS,etal.Multiband Audio Modeling for Single-channel AcousticSource Separation[C].ICASSP 2004,Montreal,Quebec,Canada,2004,5:641-644.
[33]BHIKSHA RAJ,MADHUSUDANA V S SHASHANKA,PARIS SMARAGDIS.Latent Dirichlet Decomposition forSingle Channel Speaker Separation[C].ICASSP 2006,Toulouse,France,2006,5:821-824.
[34]TUOMAS VIRTANEN.Monaural Sound Source Separationby Nonnegative Matrix Factorization with Temporal Conti-nuity and Sparseness Criteria[J].IEEE Transactions onAudio,Speech,and Language Processing,2007,15(3):1 066-1 074.
[35]C啨DRIC F啨VOTTE,SIMON J GODSILL.A Bayesian Ap-proach for Blind Separation of Sparse Sources[J].IEEETransactions on Audio,Speech,and Language Processing,2006,14(6):2 174-2 188.
[36]徐尚志,苏勇,叶中付.欠定条件下的盲分离算法[J].数据采集与处理,2006,21(2):128-132.
[37]HAROLD SZU,PORNCHAI CHANYAGORN,IVICA KO-PRIVA.Sparse Coding Blind Source Separation ThroughPowerline[J].Neurocomputing,2002,48(1):1 015-1 020.
[38]邱伟,黄秀轩.计算听觉场景分析介绍[J].高技术通讯,2002,12(1):106-110.
[39]DELIANG WANG,GUY J BROWN.Separation of Speechfrom Interfering Sounds Based on Oscillatory Correlation[J].IEEE Transactions on Neural Networks,1999,10(3):684-697.
[40]GUY J BROWN,DELIANG WANG.Separation of Speechby Computational Auditory Scene Analysis[C]//J Benes-ty,S Makino,J Chen.Speech Enhancement:What’s New?New York,USA:Springer,2005,371-402.
[41]MINGYANG WU,DELIANG WANG,GUY J BROWN.AMultipitch Tracking Algorithm for Noisy Speech[J].IEEETransactions on Speech and Audio Processing,2003,11(3):229-241.
[42]GUONING HU,DELIANG WANG.Monaural Speech Seg-regation Based on Pitch Tracking and Amplitude Modula-tion[J].IEEE Transactions on Neural Networks,2004,15(5):1 135-1 150.
[43]PENG LI,YONG GUAN,BO XU,et al.Monaural SpeechSeparation Based on Computational Auditory Scene Analy-sis and Objective Quality Assessment of Speech[J].IEEETransactions on Audio,Speech,and Language Processing,2006,14(6):2 014-2 023.
[44]MOHAMMAD HRADFAR,RICHARD MDANSEREAU,ABOLGHASEM SAYADIYAN.A Maximum LikelihoodEstimation of Vocal-tract-related Filter Characteristics forSingle Channel Speech Separation[J].EURASIP Journalon Audio,Speech,and Music Processing,2007(1):1-15.
[45]张晓冬,王桥,吴乐南.利用脊的特征进行信号盲分离[J].电子学报,2004,32(7):1 156-1 159.
[46]蔡权伟,魏平,肖先赐.基于模型拟合的重叠信号盲分离方法[J].电子学报,2005,33(10):1 794-1 798.
[47]EMMANUEL VINCENT,R GRIBONVAL,C FEVOTTE.Performance Measurement in Blind Audio Source Sepa-rateion[J].IEEE Transactions on Audio,Speech,andLanguage Processing,2006,14(4),1 462-1 469.
[48]蔡权伟,魏平,肖先赐.单信道多信号分量分离[J].通信学报,2006,27(6):49-56.
[49]CAI QUANWEI,WEI PING,XIAO XIANCI.Single Chan-nel Signal Component Separation Using Bayesian Estima-tion[J].Journal of Systems Engineering and Electronics,2007,18(1):33-39.
[50]张秀琦,刘辉,郑建斌,等.信号处理技术在重叠化学信号分析中的应用[J].化学进展,2002,14(3):174-181.
[51]ASHWIN RAO,RAMDAS KUMARESAN.Separation ofCochannel Signals Using the Parametric Constant ModulusAlgorithm[C].ICASSP 1996,Atlanta,Georgia,USA,1996,5:2 686-2 689.
[52]ARVIND V KEERTHI,JOHN J SHYNK.Separation of Co-channel Signals in TDMA Mobile Radio[J].IEEE Trans-actions on Signal Processing,1998,46(10):2 684-2 697.
[53]王正明,段晓君.信噪分离的频域方法[J].中国科学(E辑),1999,29(6):525-531.
[54]易东云,朱炬波,王正明.三频带信号分离的样条函数方法[J].电子学报,1999,27(1):54-57.
[55]周成,刘钊.基于独立分量分析的单通道语音降噪算法研究[J].激光与红外,2005,35(5):374-377.
[56]MASAKI KAWAKASTU.Application of ICAto MEG NoiseReduction[C].ICA2003,Nara,Japan,2003,535-541.
[57]OMAR A YESTE-OJEDA,JESUS GRAJAL.Cyclostation-arity-based Signal Separation in Interceptors Based on aSingle Sensor[C].2008 IEEE Radar Conference,RomeItaly,2008,169-174.
[58]SAM HEIDARI,CHRYSOSTOMOS L NIKIAS.Co-chan-nel Interference Mitigation in the Time-scale Domain:theCIMTS Algorithm[J].IEEE Transactions on Signal Pro-cessing,1996,44(9):2 151-2 162.
[59]CECILIA DI RUBERTO,ANDREW DEMPSTER,SHA-HID KHAN,et al.Segmentation of Blood Images UsingMorphological Operators[C].ICPR 2000,Barcelona,Spain,2000,3:397-400.
[60]郑宜珍,秦前清.小波包变换在重叠地震信号分离中的应用[J].计算机工程与应用,2004,40(26):176-177.
[61]HOONBIN HONG,MING LIANG.Separation of FaultFeatures from a Single-channel Mechanical Signal MixtureUsing Wavelet Decomposition[J].Mechanical Systems and Signal Processing,2007,21:2 025-2 040.
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心