Several visual feature extraction algorithms have recently appeared in the literature, with the goal of reducing the computational complexity of state-of-the-art solutions (e.g., SIFT and SURF). Therefore, it is necessary to evaluate the performance of these emerging visual descriptors in terms of processing time, repeatability and matching accuracy, and whether they can obtain competitive performance in applications such as image retrieval. This paper aims to provide an up-to-date detailed, clear, and complete evaluation of local feature detector and descriptors, focusing on the methods that were designed with complexity constraints, providing a much needed reference for researchers in this field. Our results demonstrate that recent feature extraction algorithms, e.g., BRISK and ORB, have competitive performance requiring much lower complexity and can be efficiently used in low-power devices
I. INTRODUCTIONT HANKS to the increasing availability of inexpensive small-scale hardware devices with wireless communication capabilities, distributed sensor networks are steadily gaining popularity in a wide range of application scenarios, including security, environmental monitoring and elderly care. Acoustic source localization is a feature that is often required Manuscript
Abstract. In this manuscript, we formulate the problem of source localization based on Time Differences of Arrival (TDOAs) in the TDOA space, i.e. the Euclidean space spanned by TDOA measurements. More specifically, we show that source localization can be interpreted as a denoising problem of TDOA measurements. As this denoising problem is difficult to solve in general, our analysis shows that it is possible to resort to a relaxed version of it. The solution of the relaxed problem through linear operations in the TDOA space is then discussed, and its analysis leads to a parallelism with other stateof-the-art TDOA denoising algorithms. Additionally, we extend the proposed solution also to the case where only TDOAs between few pairs of microphones within an array have been computed. The reported denoising algorithms are all analytically justified, and numerically tested thorough simulative campaign. TDOA space and TDOA denoising and TDOA redundancy and Source localization
The curse of outlier measurements in estimation problems is a well known issue in a variety of fields. Therefore, outlier removal procedures, which enables the identification of spurious measurements within a set, have been developed for many different scenarios and applications. In this paper, we propose a statistically motivated outlier removal algorithm for time differences of arrival (TDOAs), or equivalently range differences (RD), acquired at sensor arrays. The method exploits the TDOA-space formalism and works by only knowing relative sensor positions. As the proposed method is completely independent from the application for which measurements are used, it can be reliably used to identify outliers within a set of TDOA/RD measurements in different fields (e.g. acoustic source localization, sensor synchronization, radar, remote sensing, etc.). The proposed outlier removal algorithm is validated by means of synthetic simulations and real experiments. Index TermsTDOA space, TDOA measurements, range differences, outlier removal.
In this paper we discuss a method for localizing acoustic reflectors in space based on acoustic measurements on source-to-microphone reflective paths. The method converts Time of Arrival (TOA) and Time Difference of Arrival (TDOA) into quadratic constraints on the line corresponding to the reflector. In order to be robust against measurement errors we derive an exact solution for the min imization of a cost function that combines an arbitrary number of quadratic constraints. Moreover we propose a new method for the analytic prediction of reflector localization accuracy. This method is sufficiently general to be applicable to a wide range of estimation problems.Index Terms-Microphone arrays. space-time audio process ing. environment reconstruction. acoustic reflector localizationKnowing the geometry of the acoustic environment can be very use ful for numerous space-time processing applications. For example. in [1] source localization is approached using a maximum likeli hood estimator whose data model incorporates the prediction of the early reflections. In [2] an environment-aware acoustic rendering system is proposed. in which early reflections are included in the propagation model from the loudspeakers to the listening area. This has the result of making their rendering system robust to mild re verberation. Consequently. the problem of estimating the geometry of the environment through acoustic acquisitions is an area of in creasing interest. In [3] a method is proposed for estimating the reflectors based on the inverse mapping of the acoustic multi-path propagation problem. In [4] the parameters of a constrained room model are estimated through f\ -regularized least-squares. In [5] the problem of the estimation of the room geometry is approached through the measurement of the Times of Arrival (TOAs) of the re flective path from the source to the microphone. Here TOAs are converted into geometric constraints that locate the line that the re flector lies upon. For a single source-microphone pair such con straints express that this line should be tangential to an ellipse that is parameterized by the locations of the source and the microphone and by the TOA. Using multiple observations with a microphone array. the reflector is found as the common tangent to all such el lipses. which is estimated through the iterative minimization of a The cost functions defined in [5. 6] are inherently nonlinear. therefore they exhibit numerous local minima in which adaptive optimization algorithms could easily become trapped. particularly in the presence of relevant measurement errors. In this paper we propose an exact minimization procedure that determines the cor rect global minimum of the cost function while circumventing the problem of local minima. The problem is reformulated as the con strained minimization of a second-order polynomial. which admits an exact solution. This reformulation is inspired by [7]. where a source localization problem is approached with an exact minimiza tion of a constrained least-squares c...
Abstract-This work describes the implementation of an object recognition service on top of energy and resourceconstrained hardware. A complete pipeline for object recognition based on the BRISK visual features is implemented on Intel Imote2 sensor devices. The reference implementation is used to assess the performance of the object recognition pipeline in terms of processing time and recognition accuracy.
In this paper we present a visibility-based beam\ud tracing solution for the simulation of the acoustics of environment\ud that makes use of a projective geometry representation. More\ud specifically, projective geometry turns out to be useful for the precomputation\ud of the visibility among all the reflectors in the environment.\ud The simulation engine has a straightforward application\ud in the rendering of the acoustics of virtual environments using\ud loudspeaker arrays. More specifically, the acoustic wavefield is\ud conceived as a superposition of acoustic beams, whose parameters\ud (i.e. origin, orientation and aperture) are computed using the fast\ud beam tracing methodology presented here. This information is\ud processed by the rendering engine to compute spatial filters to be\ud applied to the loudspeakers within the array. Simulative results\ud show that an accurate simulation of the acoustic wavefield can\ud be obtained using this approach
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.