ICASSP99 Signal Separation

Signal Separation
Home Full List of Titles 1: Speech Processing CELP Coding Large Vocabulary Recognition Speech Analysis and Enhancement Acoustic Modeling I ASR Systems and Applications Topics in Speech Coding Speech Analysis Low Bit Rate Speech Coding I Robust Speech Recognition in Noisy Environments Speaker Recognition Acoustic Modeling II Speech Production and Synthesis Feature Extraction Robust Speech Recognition and Adaptation Low Bit Rate Speech Coding II Speech Understanding Language Modeling I 2: Speech Processing, Audio and Electroacoustics, and Neural Networks Acoustic Modeling III Lexical Issues/Search Speech Understanding and Systems Speech Analysis and Quantization Utterance Verification/Acoustic Modeling Language Modeling II Adaptation /Normalization Speech Enhancement Topics in Speaker and Language Recognition Echo Cancellation and Noise Control Coding Auditory Modeling, Hearing Aids and Applications of Signal Processing to Audio and Acoustics Spatial Audio Music Applications Application - Pattern Recognition & Speech Processing Theory & Neural Architecture Signal Separation Application - Image & Nonlinear Signal Processing 3: Signal Processing Theory & Methods I Filter Design and Structures Detection Wavelets Adaptive Filtering: Applications and Implementation Nonlinear Signals and Systems Time/Frequency and Time/Scale Analysis Signal Modeling and Representation Filterbank and Wavelet Applications Source and Signal Separation Filterbanks Emerging Applications and Fast Algorithms Frequency and Phase Estimation Spectral Analysis and Higher Order Statistics Signal Reconstruction Adaptive Filter Analysis Transforms and Statistical Estimation Markov and Bayesian Estimation and Classification 4: Signal Processing Theory & Methods II, Design and Implementation of Signal Processing Systems, Special Sessions, and Industry Technology Tracks System Identification, Equalization, and Noise Suppression Parameter Estimation Adaptive Filters: Algorithms and Performance DSP Development Tools VLSI Building Blocks DSP Architectures DSP System Design Education Recent Advances in Sampling Theory and Applications Steganography: Information Embedding, Digital Watermarking, and Data Hiding Speech Under Stress Physics-Based Signal Processing DSP Chips, Architectures and Implementations DSP Tools and Rapid Prototyping Communication Technologies Image and Video Technologies Automotive Applications / Industrial Signal Processing Speech and Audio Technologies Defense and Security Applications Biomedical Applications Voice and Media Processing Adaptive Interference Cancellation 5: Communications, Sensor Array and Multichannel Source Coding and Compression Compression and Modulation Channel Estimation and Equalization Blind Multiuser Communications Signal Processing for Communications I CDMA and Space-Time Processing Time-Varying Channels and Self-Recovering Receivers Signal Processing for Communications II Blind CDMA and Multi-Channel Equalization Multicarrier Communications Detection, Classification, Localization, and Tracking Radar and Sonar Signal Processing Array Processing: Direction Finding Array Processing Applications I Blind Identification, Separation, and Equalization Antenna Arrays for Communications Array Processing Applications II 6: Multimedia Signal Processing, Image and Multidimensional Signal Processing, Digital Signal Processing Education Multimedia Analysis and Retrieval Audio and Video Processing for Multimedia Applications Advanced Techniques in Multimedia Video Compression and Processing Image Coding Transform Techniques Restoration and Estimation Image Analysis Object Identification and Tracking Motion Estimation Medical Imaging Image and Multidimensional Signal Processing Applications I Segmentation Image and Multidimensional Signal Processing Applications II Facial Recognition and Analysis Digital Signal Processing Education Author Index A B C D E F G H I J K L M N O P Q R S T U V W X Y Z	Blind Separation of Temporomandibular Joint Sound Signals Authors: Yinchao Guo, Time-Frequency Lab, c/o Chris Koh, Division of SEM, School of Mechanical and Production Engineering, Nanyang Technological University, Singapore 639798 (Singapore) Farook Sattar, Division of Information Engineering, School of EEE, Nanyang Technological University, Nanyang Technological University, Singapore 639798 (Singapore) Christopher Koh, School of Mechanical and Production Engineering, Nanyang Technological University, Singapore 639798 (Singapore) Page (NA) Paper number 1226 Abstract: In order to develop a cheap, efficient and reliable diagnostic tool for the detection of temporomandibular joint disorders (TMD), sounds from the temporomandibular joint (TMJ) are recorded using a pair of microphone inserted in the auditory canals. However, the TMJ sounds originating from one side of head can also be picked up by microphone at the other side. Blind source separation (BSS) is thus proposed as a method to recover the original sound. The authors propose to use non-casual filters for the separation of TMJ signals. The algorithm is based on information theory and is an extension of early work by Torkkola. The separation was successful and the output can now be used for subsequent analysis of TMJ sounds. IC991226.PDF (From Author) IC991226.PDF (Rasterized) TOP Generalized Anti-Hebbian Learning For Source Separation Authors: Hsiao-Chun Wu, Jose C. Principe, Page (NA) Paper number 2394 Abstract: The information-theoretic framework for source separation is highly suitable. However the choice of the nonlinearity or the estimation of the multidimensional joint probability density function are nontrivial. We propose here a generalized Gaussian model to construct a generalized blind source separation network based on the minimum entropy principle. This new separation network can suppress the interference to a significant amount compared to the traditional LMS-echo-canceler. The simulation is given to show the disparity of the performance as a varies. Finally how to choose the appropriate a in our generalized anti-Hebbian rule is discussed. IC992394.PDF (From Author) IC992394.PDF (Rasterized) TOP Complex Independent Component Analysis By Nonlinear Generalized Hebbian Learning with Rayleigh Nonlinearity Authors: Eraldo Pomponi, Simone Fiori, Francesco Piazza, Page (NA) Paper number 1898 Abstract: The aim of this paper is to present a non-linear Extension of the Sanger's Generalized Hebbian Algorithm to the processing of complex-valued data. A possible choice of the involved non-linearity is discussed recalling the Sudjianto-Hassoun nterpretation of the non-linear Hebbian learning. Extension of this interpretation to the complex case leads to a nonlinearity called Rayleigh function, which allows for separating mixed independent complex-valued source signals. IC991898.PDF (Scanned) TOP Sparse Basis Selection, ICA, and Majorization: Towards a Unified Perspective Authors: Ken Kreutz-Delgado, Bhaskar D Rao, Page (NA) Paper number 2411 Abstract: Sparse solutions to the linear inverse problem Ax = y and the determination of an environmentally adapted overcomplete dictionary (the columns of A) depend upon the choice of a ``regularizing function'' d(x) in several recently proposed procedures. We discuss the interpretation of d(x) within a Bayesian framework, and the desirable properties that ``good'' (i.e., sparsity ensuring) regularizing functions, d(x) might have. These properties are: Schur-concavity (d(x) is consistent with majorization); concavity (d(x) has sparse minima); parameterizability (d(x) is drawn from a large, parameterizable class); and factorizability of the gradient of d(x) in a certain manner. The last property (which naturally leads one to consider separable regularizing functions) allows d(x) to be efficiently minimized subject to Ax=y using an Affine Scaling Transformation (AST)-like algorithm ``adapted'' to the choice of d(x). A Bayesian framework allows the algorithm to be interpreted as an Independent Component Analysis (ICA) procedure. IC992411.PDF (From Author) IC992411.PDF (Rasterized) TOP Two Spatio-temporal Decorrelation Learning Algorithms and their Application to Multichannel Blind Deconvolution Authors: Seungjin Choi, School of Electrical and Electronics Engineering, Chungbuk National University, KOREA (Korea) Andrzej Cichocki, Brain-style Information Systems Research Group, BSI, RIKEN, JAPAN (Japan) Shun-ichi Amari, Brain-style Information Systems Research Group, BSI, RIKEN, JAPAN (Japan) Page (NA) Paper number 1819 Abstract: In this paper we present and compare two different spatio-temporal decorrelation learning algorithms for updating the weights of a linear feedforward network with FIR synapses (MIMO FIR filter). Both standard gradient and the natural gradient are employed to derive the spatio-temporal decorrelation algorithms. These two algorithms are applied to multichannel blind deconvolution task and their performance is compared. The rigorous derivation of algorithms and computer simulation results are presented. IC991819.PDF (From Author) IC991819.PDF (Rasterized) TOP Adaptive Paraunitary Filter Banks for Principal and Minor Subspace Analysis Authors: Scott C Douglas, Dept. of Electrical Engr., Southern Methodist Univ., Dallas, TX USA (USA) Shun-ichi Amari, RIKEN Brain Science Institute, Saitama, JAPAN (Japan) S.-Y. Kung, Dept. of Electrical Engineering, Princeton Univ., Princeton, NJ USA (USA) Page (NA) Paper number 1786 Abstract: Paraunitary filter banks are important for several signal processing tasks. In this paper, we consider the task of adapting the coefficients of a multichannel FIR paraunitary filter bank via gradient ascent or descent on a chosen cost function. The proposed generalized algorithms inherently adapt the system's parameters in the space of paraunitary filters. Modifications and simplifications of the techniques for spatio-temporal principal and minor subspace analysis are described. Simulations verify one algorithm's useful behavior in this task. IC991786.PDF (From Author) IC991786.PDF (Rasterized) TOP