The LOCATA Challenge: Acoustic Source Localization and Tracking

Abstract

The ability to localize and track acoustic events is a fundamental prerequisite for equipping machines with the ability to be aware of and engage with humans in their surrounding environment. However, in realistic scenarios, audio signals are adversely affected by reverberation, noise, interference, and periods of speech inactivity. In dynamic scenarios, where the sources and microphone platforms may be moving, the signals are additionally affected by variations in the source-sensor geometries. In practice, approaches to sound source localization and tracking are often impeded by missing estimates of active sources, estimation errors, as well as false estimates. The aim of the LOCAlization and TrAcking (LOCATA) Challenge is an open-access framework for the objective evaluation and benchmarking of broad classes of algorithms for sound source localization and tracking. This article provides a review of relevant localization and tracking algorithms and, within the context of the existing literature, a detailed evaluation and dissemination of the LOCATA submissions. The evaluation highlights achievements in the field, open challenges, and identifies potential future directions.

Sparse Bayesian Learning for Acoustic Source Localization

“…4. Due to rotations of the eigenmike within the shockmount, it is highly sensitive to scattering effects [17] which results in high azimuth error (Fig. 4, note that the two plots have different vertical range).…”

“…The LOCATA [15] ambient noise (from a road in front of the building). Further details and assumptions about the LOCATA dataset can be found in [15][16][17]. We consider a 12-microphone pseudospherical array named robot-head, 32-microphone spherical array named eigenmike and a 15-microphone non-uniform array named dicit.…”

“…The IEEE-AASP Challenge on sound source localization and tracking (LOCATA [15,16]) provides an open-access data corpus of indoor multi-channel audio recordings in presence of multiple mobile sources and their ground truth for performance evaluation. Since its release, various methods for localization have been applied on this dataset (see review [17]. None of the methods have explored CS based processing for DOA estimation.…”

Evaluation and Comparison of Three Source Direction-of-Arrival Estimators Using Relative Harmonic Coefficients

“…In the past decades, source direction-of-arrival (DOA) estimation [1,2] has been extensively investigated in the research community since it is an essential component in many spatial signal processing techniques and applications including source dereverberation, speech separation [3], automatic speech recognition [4] and automated camera steering [5].…”

Overlapping Speaker Segmentation Using Multiple Hypothesis Tracking of Fundamental Frequency

“…Diarization is performed without any prior knowledge of the number of speakers or the amount of speech the recording contains. Accurate diarization has become increasingly important in recent years for a multitude of tasks including voice control of smart devices and robot audition [2], [3]. Diarization is also required for applications such as speaker indexing [4], automatic speech recognition (ASR) [5] and enabling the use of single speaker-based algorithms in multi-speaker domains [6].…”