Independent Component Analysis.

Binary time-frequency masks are powerful tools for the separation of sources from a single mixture. Perfect demixing via binary time-frequency masks is possible provided the time-frequency representations of the sources do not overlap: a condition we call W-disjoint orthogonality. We introduce here the concept of approximate W-disjoint orthogonality and present experimental results demonstrating the level of approximate W-disjoint orthogonality of speech in mixtures of various orders. The results demonstrate that there exist ideal binary time-frequency masks that can separate several speech signals from one mixture. While determining these masks blindly from just one mixture is an open problem, we show that we can approximate the ideal masks in the case where two anechoic mixtures are provided. Motivated by the maximum likelihood mixing parameter estimators, we define a power weighted two-dimensional (2-D) histogram constructed from the ratio of the time-frequency representations of the mixtures that is shown to have one peak for each source with peak location corresponding to the relative attenuation and delay mixing parameters. The histogram is used to create time-frequency masks that partition one of the mixtures into the original sources. Experimental results on speech mixtures verify the technique. Example demixing results can be found online at http://alum.mit.edu/www/rickard/bss.html.

Blind separation of speech mixtures via time-frequency masking

Active noise control exploits the long wavelengths associated with low frequency sound. It works on the principle of destructive interference between the sound fields generated by the original primary sound source and that due to other secondary sources, acoustic outputs of which can be controlled. The acoustic objectives of different active noise control systems and the electrical control methodologies that are used to achieve these objectives are examined. The importance of having a clear understanding of the principles behind both the acoustics and the electrical control in order to appreciate the advantages and limitations of active noise control is emphasized. A brief discussion of the physical basis of active sound control that concentrates on three-dimensional sound fields is presented. >

Active noise control

On typical echo paths, the proportionate normalized least-mean-squares (PNLMS) adaptation algorithm converges significantly faster than the normalized least-mean-squares (NLMS) algorithm generally used in echo cancelers to date. In PNLMS adaptation, the adaptation gain at each tap position varies from position to position and is roughly proportional at each tap position to the absolute value of the current tap weight estimate. The total adaptation gain being distributed over the taps is carefully monitored and controlled so as to hold the adaptation quality (misadjustment noise) constant. PNLMS adaptation only entails a modest increase in computational complexity.

Proportionate normalized least-mean-squares adaptation in echo cancelers

There has been a great deal of material in the area of discrete-time transforms that has been published in recent years. This book does an excellent job of presenting important aspects of such material in a clear manner. The book has 11 chapters and a very useful appendix. Seven of these chapters are essentially devoted to the Fourier series/transform, discrete Fourier transform, fast Fourier transform (FFT), and applications of the FFT in the area of spectral estimation. Chapters 8 through 10 deal with many other discrete-time transforms and algorithms to compute them. Of these transforms, the KarhunenLoeve, the discrete cosine, and the Walsh-Hadamard transform are perhaps the most well-known. A lucid discussion of number theoretic transforms i5 presented in Chapter 11. This reviewer feels that the authors have done a fine job of compiling the pertinent material and presenting it in a concise and clear manner. There are a number of problems at the end of each chapter, an appreciable number of which are challenging. The authors have included a comprehensive set of references at the end of the book. In brief, this book is a valuable contribution to the general areas of signal processing and communications. It can be used for a graduate level course in perhaps two ways. One would be to cover the first seven chapters in great detail. The other would be to cover the whole book by focussing on different topics in a selective manner. This book  by  Elliott  and Rao  is extremely  useful to researchers/engineers who are working  in the areas of signal processing and communications. It i s also an excellent reference book, and hence a valuable addition to one’s library

http://ieeexplore.ieee.org/iel5/5/31347/01457444.pdf

Multirate digital signal processing

Sweep sounds perceived in a rectangular parallelepiped reverberation room.

PURPOSE:To achieve the decrease in amount of operations and the like by using a plurality of measured sound transfer characteristics, estimating the physical pole of a object acoustic system, using the estimated pole as the fixed amount, and simulating the sound transfer characteristics in the acoustic system. CONSTITUTION:A sound source 51 and a sound receiver 52 are arranged in a space 11 in a chamber which is an object. The sound transfer characteristic between the sound source 51 and the sound receiver 52 is measured with a sound-transfer-characteristic measuring part 53 using these parts. A plurality of the measured sound transfer characteristics are sent into a pole estimating part 54, and the physical pole is estimated based on the sound transfer characteristics. The pole which is estimated in the estimating part 54 beforehand is supplied into a fixed AR filter 55. The pole simulating operation of the actual sound transfer characteristics is performed with the filter 55 and a variable MA filter 56 which is provided at the front stage or the rear stage of the filter 55. When the unknown sound transfer characteristics are estimated and applied on a simulating apparatus, the number of parameters required for the estimation can be reduced, and the reduction in amount of operation and the improvement of the estimating speed can be achieved.

Method for simulating sound transfer characteristic

The author describes the principle and performance of AMNOR, which consists of a compact microphone array and signal processing techniques that include an adaptive filter. It achieves high-SNR (signal-to-noise) sound reception, which is essential in areas such as teleconference systems and speech-recognition systems. The basic distinction of AMNOR compared to previous designs is that it realizes higher SNR by accepting some degree of degradation of a desired speech signal. A noise power minimization problem under second-order constraints is solved in the AMNOR system using an adaptive filter. An experiment conducted in an ordinary room showed that AMNOR realized a 21-dB SNR improvement, while a conventional directional microphone attained a 3-dB SNR improvement. >

Adaptive microphone array system for noise reduction (AMNOR) and its performance studies

PROBLEM TO BE SOLVED: To suppress an echo even in the case that an initial reach channel is unknown. SOLUTION: A de-reverberation apparatus includes: a delay adding unit 41 for generating a delay added signal delayed at least one of a plurality of acoustic signals for only a predetermined delay time; and a de-reverberation processing unit 23 j which performs de-reverberation processing using the delay added signal. Thus, by adding a delay to an input signal other than a representative channel, the predetermined representative channel can be set as a channel that the acoustic signal first reaches. COPYRIGHT: (C)2010,JPO&INPIT

De-reverberation apparatus and de-reverberation method

Graduate School of Engineering, Tokyo Denki University,5 Senju-Asahi-cho, Adachi-ku, Tokyo, 120–8551 Japan(Received 10 October 2014, Accepted for publication 21 November 2014)Keywords: Reverberation time, Impulse response measurement, Swept sine, Signal-to-noise ratioPACS number: 43.58. e, 45.58.Gn, 43.60. c [doi:10.1250/ast.36.344]1. IntroductionReverberation time is a basic characteristic of roomacoustics, which is calculated from the impulse response (IR)of the room. In the measurement of reverberation time fromthe IR, a high signal-to-noise ratio (SNR) is required overa wide frequency band. In a previous paper, the authorsproposed the use of a constant signal-to-noise ratio sweptsine (CSN-SS) signal that provides a constant SNR of themeasured IR over the frequency band of interest [1,2]. In thispaper, the authors propose an eﬀective measurement methodfor reverberation time using the CSN-SS signal.2. Control of SNR using CSN-SS signalFigure 1 shows the principle of IR measurement in thefrequency domain. HðkÞ is the transfer function of a targetsystem, which is the Fourier transform of IR, where k is thediscrete frequency number, although it is omitted in Fig. 1 forsimplicity. HðkÞ can be measured using a measurement signalSðkÞ, such as a swept sine signal. The noise componentexpressed as NðkÞ=SðkÞ is included in the measurement result,where NðkÞ is an environmental noise added to the systemoutput.The SNR of this result is the power ratio of HðkÞ toNðkÞ=SðkÞ:SNRðkÞ¼jHðkÞj

https://www.jstage.jst.go.jp/article/ast/36/4/36_E1472/_pdf

Yutaka Kaneda

Papers

Sweep sounds perceived in a rectangular parallelepiped reverberation room.

Method for simulating sound transfer characteristic

Adaptive microphone array system for noise reduction (AMNOR) and its performance studies

De-reverberation apparatus and de-reverberation method

Effective measurement method for reverberation time using a constant signal-to-noise ratio swept sine signal