A Compressed Encoding Scheme for Approximate Tdoa Estimation

EURASIP Journal on Advances in Signal Processing, 2007

Speaker localization with microphone arrays has received significant attention in the past decade as a means for automated speaker tracking of individuals in a closed space for videoconferencing systems, directed speech capture systems, and surveillance systems. Traditional techniques are based on estimating the relative time difference of arrivals (TDOA) between different channels, by utilizing crosscorrelation function. As we show in the context of speaker localization, these estimates yield poor results, due to the joint effect of reverberation and the directivity of sound sources. In this paper, we present a novel method that utilizes a priori acoustic information of the monitored region, which makes it possible to localize directional sound sources by taking the effect of reverberation into account. The proposed method shows significant improvement of performance compared with traditional methods in "noise-free" condition. Further work is required to extend its capabilities to noisy environments.

2008

Abstract In this paper we present a novel approach to directly recover the location of both microphones and sound sources from time-difference-of-arrival measurements only. No approximation solution is required for initialization and in the absence of noise our approach is guaranteed to always recover the exact solution. Our approach only requires solving linear equations and matrix factorization. We demonstrate the feasibility of our approach with synthetic data.

EURASIP Journal on Advances in Signal Processing, 2011

TDOA-(time difference of arrival-) based algorithms are common methods for speech source localization. The generalized cross correlation (GCC) method is the most important approach for estimating TDOA between microphone pairs. The performance of this method significantly degrades in the presence of noise and reverberation. This paper addresses the problem of 3D localization in joint noisy and reverberant conditions and a single-speaker scenario. We first propose a modification to make the GCC-PHAse transform (GCC-PHAT) method robust against environment noise. Then, we use an iterative technique that employs location estimation to improve TDOAs accuracy. Extensive experiments on both simulated and real (practical) data (in a single-source scenario) show the capability of the proposed methods to significantly improve TDOA accuracy and, consequently, source location estimates.

2006 International Workshop on Intelligent Solutions in Embedded Systems, 2006

A new acoustic localization scheme is proposed, which can be applied in sensor networks with low communication bandwidth. The first step of the two-stage algorithm is distributed and is evaluated at the sensors having possibly low communication capabilities. The sensors detect various events in the acoustic signal and transmit only the event descriptors to the base station where the sensor fusion is calculated in the second step using a novel generalized consistency function. Test results validate the performance of the proposed system in a noisy and reverberant environment. According to experiments, the proposed system decreases the necessary communication bandwidth with multiple orders of magnitude and still provides high accuracy comparable with that of the sophisticated beam-forming methods.

Anais de XXXVII Simpósio Brasileiro de Telecomunicações e Processamento de Sinais, 2019

In this paper we propose a new time difference delay of arrival technique based on the semblance multichannel coherency function for the problem of sound source localization. The proposed algorithm was tested on recordings from an Unmanned Aerial Vehicle (UAV) equipped with an array of 8 microphones, for estimating the azimuth and elevation angles of a speech based source. Our results shown that the semblance method has proven to have a good performance, obtaining good results regardless of the ego noise even in cases where the signalto-noise ratio (SNR) was very low.

Multidimensional Systems and Signal Processing, 2013

In this paper we consider a class of geometric methods for acoustic source localization based on range differences (or time differences of arrival), and we offer a new and unifying perspective on such methods based on the adoption of a multidimensional reference frame obtained by adding the range difference coordinate to the spatial coordinates of the source. In this extended coordinate system the working principles of a wide range of source localization methods becomes clear and immediate. The space-range reference frame, however, has a practical purpose as well, as it can be used for gaining insight on why some configurations of microphones lead to better localization performance than others and it suggests methods for improving existing localization techniques. In particular, we derive a closed-form solution of the constrained least-squares localization problem for linear arrays of microphones.

Proceedings of the Detection and Classification of Acoustic Scenes and Events 2019 Workshop (DCASE2019), 2019

In this work, we show a simultaneous sound event localization and detection (SELD) system, with enhanced acoustic features, in which we propose using the well-known Generalized Cross Correlation (GCC) PATH algorithm, to augment the magnitude and phase regular Fourier spectra features at each frame. GCC-PHAT has already been used for some time to calculate the Time Difference of Arrival (TDoA) in simultaneous audio signals, in moderately reverberant environments, using classic signal processing techniques, and can assist audio source localization in current deep learning machines. The neural net architecture we used is a Convolutional Recurrent Neural Network (CRNN), and is tested using the sound database prepared for the Task 3 of the 2019 DCASE Challenge. In the challenge results, our proposed system was able to achieve 20.8 of direction of arrival error, 85.6% frame recall, 86.5% F-score and 0.22 error rate detection in evaluation samples.

A computer-based algorithm that localizes sounds in near-real time has been developed. The algorithm takes input from two microphones and estimates the position of the sound source relative to the microphone array. The algorithm requires no a priori knowledge of the stimuli to be localized. The accuracy of the algorithm was tested using binaural recordings from a pair of microphones mounted in the ear canals of an acoustic mannequin. Sounds were played at 5 degree steps around the mannequin and the outputs were recorded at the entrance to each ear canal. These recordings were fed into the algorithm that estimated the location of the incoming sound on the horizontal plane. The algorithm utilizes a Head-Related Transfer Function (HRTF) to estimate the location of incoming sound stimuli. Although the HRTF of the acoustic mannequin was used, any HRTF can be inserted into the algorithm, allowing for predictions about individual performance differences. The results of this effort have been highly encouraging: the algorithm was able to identify accurately the location of a variety of sounds, committing an average of 2.9 degrees of unsigned localization error. Better than chance performance was found in noisy conditions of up to a -10 dB signal-tonoise ratio. The initial purpose of this algorithm is to predict the localization performance afforded by different types of combat helmets. Current and future encapsulating helmet designs are likely to impede localization performance, and an accurate localization model would be an invaluable tool in the helmet selection process. Adapting the model for use as a highly accurate machine-based localizer is an additional goal of this line of research. Applications for this technology include target tracking on unmanned vehicles, sniper detection, and machine-assisted sound localization.

2008 Hands-Free Speech Communication and Microphone Arrays, 2008

Comparing the different sound source localization techniques, proposed in the literature during the last decade, represents a relevant topic in order to establish advantages and disadvantages of a given approach in a real-time implementation. Traditionally, algorithms for sound source localization rely on an estimation of Time Difference of Arrival (TDOA) at microphone pairs through GCC-PHAT. When several microphone pairs are available the source position can be estimated as the point in space that best fits the set of TDOA measurements by applying Global Coherence Field (GCF), also known as SRP-PHAT, or Oriented Global Coherence Field (OGCF). A first interesting analysis compares the performance of GCF and OGCF to a suboptimal LS search method. In a second step, Adaptive Eigenvalue Decomposition is implemented as an alternative to GCC-PHAT in TDOA estimation. Comparative experiments are conducted on signals acquired by a linear array during WOZ experiments in an interactive-TV scenario. Changes in performance according to different SNR levels are reported.

Localization of sound sources in adverse environments is an important challenge in robot audition. The target sound source is often corrupted by coherent broadband noise, which introduces localization ambiguities as noise is often mistaken as the target source. To discriminate the time difference of arrival (TDOA) parameters of the target source and noise, this paper presents a binary mask for weighted generalized cross-correlation with phase transform (GCC-PHAT). Simulation and experiments on a mobile robot suggest that the proposed technique improves TDOA discrimination. It also brings the additional benefit of modulating the computing load requirement according to voice activity.