Estimation of acoustic echoes using expectation-maximization methods

Abstract

Estimation problems like room geometry estimation and localization of acoustic reflectors are of great interest and importance in robot and drone audition. Several methods for tackling these problems exist, but most of them rely on information about times-of-arrival (TOAs) of the acoustic echoes. These need to be estimated in practice, which is a difficult problem in itself, especially in robot applications which are characterized by high ego-noise. Moreover, even if TOAs are successfully extracted, the difficult problem of echolabeling needs to be solved. In this paper, we propose multiple expectation-maximization (EM) methods, for jointly estimating the TOAs and directions-of-arrival (DOA) of the echoes, with a uniform circular array (UCA) and a loudspeaker in its center for probing the environment. The different methods are derived to be optimal under different noise conditions. The experimental results show that the proposed methods outperform existing methods in terms of estimation accuracy in noisy conditions. For example, it can provide accurate estimates at SNR of 10 dB lower compared to TOA extraction from room impulse responses, which is often used. Furthermore, the results confirm that the proposed methods can account for scenarios with colored noise or faulty microphones. Finally, we show the applicability of the proposed methods in mapping of an indoor environment.

dEchorate: a calibrated room impulse response dataset for echo-aware signal processing

“…Several methods have been proposed which take into account different levels of prior information and noise (see [30,81] for a review). When the echoes' TOA and their labeling are known for 4 non-coplanar microphones, one can perform this task using geometrical reasoning as in [28,[82][83][84]. In details, the 3D coordinates of each image source can be retrieved solving a multilateration problem [85], namely the extension of the trilateration problem to 3D space, where the goal is to estimate the relative position of an object based on the measurement of its distance with respect to anchor points.…”

Acoustic Echo Sensing for Robot Localization in Buried Pipe Networks

“…The estimation the impulse response in the presence of noise can be improved [25], [26] by using a maximum likelihood (ML) (or expectation maximization (EM), or nonlinear leastsquares (NLS)) optimization method [27]. This method has been applied to localization using a single microphone [28] and an array of microphones, and the effects of signal to noise ratio, hardware transfer function [29], correlated and coloured noise, and faulty microphones [25] have been considered. Recent work has used acoustic interference to localize walls close to a small, resource constrained robot [30], showing precise localization at a distance up to around 0.5 metres.…”

“…When the robot is between any two features in the environment, there will always be this static echo which travels a total distance equal to twice the distance between those two features. In this case this is given by 25,8,18,10,20,17,12 While this description of the set of components is complicated mathematically, it is simple to implement algorithmically.…”

Detecting Acoustic Reflectors Using A Robot’s Ego-Noise

“…Here, we postulate that the acoustic structure of the ego-noise naturally produced by a robot carries enough information that may help in detecting and localizing acoustic reflectors solely based on audio recordings. In our previous works [18,19,20], we proposed an active/intrusive approach where a loudspeaker emitting a known broadband signal was attached to a drone in order to probe the environment using times of arrival. In contrast, in this work, we propose removing the loudspeaker from the setup and develop a method solely utilizing the drone's ego-noise to detect an acoustic reflector with a single microphone.…”

“…Throughout this study, we assume that the direct-path component of the egonoise within the microphone signal is known. A number of techniques could be envisioned to estimate it, e.g., dictionary learning or model-based methods calibrated using prior measurements in an anechoic chamber, e.g., [5,18,19,20], or using close-range microphones placed next to the ego-noise sources as references. This aspect is beyond the scope of this paper and is left for future iterations of this research.…”