TDOA Matrices: Algebraic Properties and Their Application to Robust Denoising With Missing Data

Measuring the time delay of arrival (TDOA) between a set of sensors is the basic setup for many applications, such as localization or signal beamforming. This paper presents the set of TDOA matrices, which are built from noise-free TDOA measurements, not requiring knowledge of the sensor array geometry. We prove that TDOA matrices are rank-two and have a special singular value decomposition decomposition that leads to a compact linear parametric representation. Properties of TDOA matrices are applied in this paper to perform denoising, by finding the TDOA matrix closest to the matrix composed with noisy measurements. This paper shows that this problem admits a closed-form solution for TDOA measurements contaminated with Gaussian noise that extends to the case of having missing data. This paper also proposes a novel robust denoising method resistant to outliers, missing data and inspired in recent advances in robust low-rank estimation. Experiments in synthetic and real datasets show significant improvements of the proposed denoising algorithms in TDOA-based localization, both in terms of TDOA accuracy estimation and localization error.

[1]  Bin Yang,et al.  Disambiguation of TDOA Estimation for Multiple Sources in Reverberant Environments , 2008, IEEE Transactions on Audio, Speech, and Language Processing.

[2]  Na Zhu,et al.  Locating and extracting acoustic and neural signals , 2011 .

[3]  Yi Ma,et al.  Robust principal component analysis? , 2009, JACM.

[4]  Jacob Benesty,et al.  Time Delay Estimation in Room Acoustic Environments: An Overview , 2006, EURASIP J. Adv. Signal Process..

[5]  Yu Hen Hu,et al.  Maximum likelihood multiple-source localization using acoustic energy measurements with wireless sensor networks , 2005, IEEE Transactions on Signal Processing.

[6]  Yubin Kuang,et al.  Stratified sensor network self-calibration from TDOA measurements , 2013, 21st European Signal Processing Conference (EUSIPCO 2013).

[7]  Zhengyou Zhang,et al.  Why does PHAT work well in lownoise, reverberative environments? , 2008, 2008 IEEE International Conference on Acoustics, Speech and Signal Processing.

[8]  Julius O. Smith,et al.  Closed-form least-squares source location estimation from range-difference measurements , 1987, IEEE Trans. Acoust. Speech Signal Process..

[9]  Geert Leus,et al.  Sparsity-Aware Multi-Source TDOA Localization , 2013, IEEE Transactions on Signal Processing.

[10]  Michael S. Brandstein,et al.  Microphone Arrays - Signal Processing Techniques and Applications , 2001, Microphone Arrays.

[11]  Darren Moore,et al.  The IDIAP Smart Meeting Room , 2002 .

[12]  Yang Weng,et al.  Total Least Squares Method for Robust Source Localization in Sensor Networks Using TDOA Measurements , 2011, Int. J. Distributed Sens. Networks.

[13]  Jean-Marc Odobez,et al.  AV16.3: An Audio-Visual Corpus for Speaker Localization and Tracking , 2004, MLMI.

[14]  Baoqi Huang,et al.  TDOA-Based Source Localization With Distance-Dependent Noises , 2015, IEEE Transactions on Wireless Communications.

[15]  Benesty,et al.  Adaptive eigenvalue decomposition algorithm for passive acoustic source localization , 2000, The Journal of the Acoustical Society of America.

[16]  Radu Horaud,et al.  A Geometric Approach to Sound Source Localization from Time-Delay Estimates , 2013, IEEE/ACM Transactions on Audio, Speech, and Language Processing.

[17]  Michael S. Brandstein,et al.  Robust Localization in Reverberant Rooms , 2001, Microphone Arrays.

[18]  Patrick A. Naylor,et al.  TDOA-Based Speed of Sound Estimation for Air Temperature and Room Geometry Inference , 2013, IEEE Transactions on Audio, Speech, and Language Processing.

[19]  Augusto Sarti,et al.  Source localization and denoising: a perspective from the TDOA space , 2015, Multidimens. Syst. Signal Process..

[20]  G. Carter,et al.  The generalized correlation method for estimation of time delay , 1976 .

[21]  C. Eckart,et al.  The approximation of one matrix by another of lower rank , 1936 .

[22]  Steven A. Tretter,et al.  Optimum processing for delay-vector estimation in passive signal arrays , 1973, IEEE Trans. Inf. Theory.

[23]  Javier Macías Guarasa,et al.  Proposal and validation of an analytical generative model of SRP-PHAT power maps in reverberant scenarios , 2016, Signal Process..

[24]  G. Carter,et al.  Special issue on time delay estimation , 1980 .

[25]  K. C. Ho,et al.  A simple and efficient estimator for hyperbolic location , 1994, IEEE Trans. Signal Process..

[26]  Arun Ross,et al.  Microphone Arrays , 2009, Encyclopedia of Biometrics.

[27]  A. Sarti,et al.  A comprehensive analysis of the geometry of TDOA maps in localization problems , 2014, ArXiv.

[28]  Dong Liu,et al.  Robust late fusion with rank minimization , 2012, 2012 IEEE Conference on Computer Vision and Pattern Recognition.

[29]  Emmanuel J. Candès,et al.  The Power of Convex Relaxation: Near-Optimal Matrix Completion , 2009, IEEE Transactions on Information Theory.

[30]  Robert D. Nowak,et al.  Transduction with Matrix Completion: Three Birds with One Stone , 2010, NIPS.

[31]  Marc Pollefeys,et al.  Direct computation of sound and microphone locations from time-difference-of-arrival data , 2008, 2008 IEEE International Conference on Acoustics, Speech and Signal Processing.

[32]  Bin Yang,et al.  A Theoretical Analysis of 2D Sensor Arrays for TDOA Based Localization , 2006, 2006 IEEE International Conference on Acoustics Speech and Signal Processing Proceedings.

[33]  K. C. Ho,et al.  A new constrained weighted least squares algorithm for TDOA-based localization , 2013, Signal Process..

[34]  K. C. Ho.  Bias Reduction for an Explicit Solution of Source Localization Using TDOA , 2012, IEEE Transactions on Signal Processing.

[35]  Dacheng Tao,et al.  GoDec: Randomized Lowrank & Sparse Matrix Decomposition in Noisy Case , 2011, ICML.

[36]  Augusto Sarti,et al.  TDOA denoising for acoustic source localization , 2015, ArXiv.

[37]  Philippe Forster,et al.  Unconditional Maximum Likelihood Performance at Finite Number of Samples and High Signal-to-Noise Ratio , 2007, IEEE Transactions on Signal Processing.

[38]  J. Odobez,et al.  AV 16 . 3 : An Audio-Visual Corpus for Speaker Localization and Tracking , .

[39]  Maurice Charbit,et al.  Slowness estimation from noisy time delays observed on non-planar arrays , 2014 .

[40]  Xavier Anguera Miró,et al.  Acoustic Beamforming for Speaker Diarization of Meetings , 2007, IEEE Transactions on Audio, Speech, and Language Processing.

[41]  R. Schmidt.  Least squares range difference location , 1996, IEEE Transactions on Aerospace and Electronic Systems.

[42]  Michael S. Brandstein,et al.  A practical methodology for speech source localization with microphone arrays , 1997, Comput. Speech Lang..

[43]  Frankie K. W. Chan,et al.  Closed-Form Formulae for Time-Difference-of-Arrival Estimation , 2008, IEEE Transactions on Signal Processing.

[44]  L. Mirsky.  SYMMETRIC GAUGE FUNCTIONS AND UNITARILY INVARIANT NORMS , 1960 .

[45]  Harvey F. Silverman,et al.  A Linear Closed-Form Algorithm for Source Localization From Time-Differences of Arrival , 2008, IEEE Signal Processing Letters.