Direction-of-arrival estimation using acoustic vector sensors in the presence of noise

Levin, Dovid Y.; Gannot, Sharon; Habets, Emanuël A. P.

doi:10.1109/icassp.2011.5946339

Cited by 32 publications

(20 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It can be shown 20 that the target function (term in curly braces) of Eq. (21) can be cast into an equivalent form as follows:…”

Section: A Doa Estimation By Steering To Direction Of Maximum Srpmentioning

confidence: 99%

“…Neither of the algorithms achieves the CRLB. In a previous study, 20 a DOA estimation method based on a search for the direction of maximum steered response power (SRP) was proposed by the authors (similar to that of Ref. 21 in the context of conventional arrays).…”

Section: Introductionmentioning

confidence: 99%

“…Maximization of SRP for a single vector-sensor can be performed with a computationally inexpensive algorithm. 20 Although the maximum SRP method with an arbitrary selection of beampattern parameter is suboptimal, the specific realization relating to ML estimation corresponds to an optimal selection of beampattern and attains the CRLB. It is further demonstrated that the minimum power distortionless response (MPDR) beamformer (from the field of signal enhancement and spectral estimation) shares the same beampattern.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Maximum likelihood estimation of direction of arrival using an acoustic vector-sensor

Levin

Habets²,

Gannot³

2012

The Journal of the Acoustical Society of America

Self Cite

View full text Add to dashboard Cite

A vector-sensor consisting of a monopole sensor collocated with orthogonally oriented dipole sensors is used for direction of arrival (DOA) estimation in the presence of an isotropic noise-field or internal device noise. A maximum likelihood (ML) DOA estimator is derived and subsequently shown to be a special case of DOA estimation by means of a search for the direction of maximum steered response power (SRP). The problem of SRP maximization with respect to a vector-sensor can be solved with a computationally inexpensive algorithm. The ML estimator achieves asymptotic efficiency and thus outperforms existing estimators with respect to the mean square angular error (MSAE) measure. The beampattern associated with the ML estimator is shown to be identical to that used by the minimum power distortionless response beamformer for the purpose of signal enhancement.

show abstract

“…It can be shown 20 that the target function (term in curly braces) of Eq. (21) can be cast into an equivalent form as follows:…”

Section: A Doa Estimation By Steering To Direction Of Maximum Srpmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Maximum likelihood estimation of direction of arrival using an acoustic vector-sensor

Levin

Habets²,

Gannot³

2012

The Journal of the Acoustical Society of America

Self Cite

View full text Add to dashboard Cite

show abstract

“…Though significant progress has been made, DOA estimation under noisy and reverberant environment is still a challenging task. To improve DOA performance, maximum steered response power (MSRP) [8] and maximum likelihood (ML) [9] methods were introduced, and the effects of noise and reverberation on these DOA estimators have been analyzed.…”

Section: Introductionmentioning

confidence: 99%

Joint Noise and Reverberation Adaptive Learning for Robust Speaker DOA Estimation with an Acoustic Vector Sensor

Wang¹,

Zou²

2018

Interspeech 2018

View full text Add to dashboard Cite

Deep neural network (DNN) based DOA estimation (DNN-DOAest) methods report superior performance but the degradation is observed under stronger additive noise and room reverberation conditions. Motivated by our previous work with an acoustic vector sensor (AVS) and the great success of DNN based speech denoising and dereverberation (DNN-SDD), a unified DNN framework for robust DOA estimation task is thoroughly investigated in this paper. First, a novel DOA cue termed as sub-band inter-sensor data ratio (Sb-ISDR) is proposed to efficiently represent DOA information for training a DNN-DOAest model. Second, a speech-aware DNN-SDD is presented, where coherence vectors denoting the probability of time-frequency points dominated by speech signals are used as additional input to facilitate the training to predict complex ideal ratio masks. Last, by stacking the DNN-DOAest on the DNN-SDD with a joint part, the unified network is jointly finetuned, which enables DNN-SDD to serve as a pre-processing front-end to adaptively generate 'clean' speech features that are easier to be correctly classified by the following DNN-DOAest for robust DOA estimation. Experimental results on simulated and recorded data confirm the effectiveness and superiority of our proposed methods under different noise and reverberations compared with baseline methods.

show abstract

“…The corresponding azimuth/elevation direction-of-arrival estimates' Cramér-Rao lower bounds have been derived in [3]- [6], [8], [11]- [14], [17]- [19], [28], [31], [32], [34], [37] for an ideal acoustic vector sensor with nominal gain/phase responses exactly as in (1); and Cramér-Rao bound curves (but unaccompanied by any formula) have been plotted in [15], [20], [21], [23], [26], [29], [38]. However, all above Cramér-Rao bound results unrealistically assume no non-ideality in the acoustic vector sensor's gain/phase response.…”

mentioning

confidence: 99%

An Hybrid Cramér-Rao Bound in Closed Form for Direction-of-Arrival Estimation by an “Acoustic Vector Sensor” With Gain-Phase Uncertainties

Tam

Wong

Song

2014

IEEE Trans. Signal Process.

View full text Add to dashboard Cite

An "acoustic vector sensor" (also known as a "vector hydrophone" in underwater or sea-surface applications) is composed of three orthogonally oriented uni-axial particle-velocity sensors, plus a "pressure-sensor" (i.e., a microphone or a hydrophone)-all collocated in a point-like spatial geometry. (This collocated setup is versatile for direction finding, because its azimuth-elevation spatial response is independent of frequency.) This paper investigates how the acoustic vector sensor's direction finding accuracy would be degraded by random deviations from its nominal gain response and/or phase response. Each type of deviation is statistically modeled herein as a random variable with a small variance, reasonably so for a well-built acoustic vector sensor. The resulting hybrid Cramér-Rao bound (HCRB) is derived exactly in open form for azimuth-elevation arrival-angle estimation, but also approximated to produce a closed form that is simple enough to yield qualitative observations. This closed-form hybrid Cramér-Rao lower bound's tightness is illustrated by a new estimator developed in this paper.Index Terms-Acoustic signal processing, acoustic velocity measurement, array signal processing, direction of arrival estimation, sonar arrays, sonar signal processing, underwater acoustic arrays.

show abstract

Direction-of-arrival estimation using acoustic vector sensors in the presence of noise

Cited by 32 publications

References 6 publications

Maximum likelihood estimation of direction of arrival using an acoustic vector-sensor

Maximum likelihood estimation of direction of arrival using an acoustic vector-sensor

Joint Noise and Reverberation Adaptive Learning for Robust Speaker DOA Estimation with an Acoustic Vector Sensor

An Hybrid Cramér-Rao Bound in Closed Form for Direction-of-Arrival Estimation by an “Acoustic Vector Sensor” With Gain-Phase Uncertainties

Contact Info

Product

Resources

About