DREGON: Dataset and Methods for UAV-Embedded Sound Source Localization

Abstract

This paper introduces DREGON, a novel publiclyavailable dataset that aims at pushing research in sound source localization using a microphone array embedded in an unmanned aerial vehicle (UAV). The dataset contains both clean and noisy in-flight audio recordings continuously annotated with the 3D position of the target sound source using an accurate motion capture system. In addition, various signals of interests are available such as the rotational speed of individual rotors and inertial measurements at all time. Besides introducing the dataset, this paper sheds light on the specific properties, challenges and opportunities brought by the emerging task of UAV-embedded sound source localization. Several baseline methods are evaluated and compared on the dataset, with real-time applicability in mind. Very promising results are obtained for the localization of a broad-band source in loud noise conditions, while speech localization remains a challenge under extreme noise levels.

Audio-Based Search and Rescue With a Drone: Highlights From the IEEE Signal Processing Cup 2019 Student Competition [SP Competitions]

“…Martin Strauss (martin.strauss@fau.de) is currently a master student with the Friedrich-Alexander University of Erlangen-Nrnberg (Germany). He is the main author of the DREGON dataset [7] which served as a basis for the competition and participated in the early design of the tasks.…”

“…Drones have already been used by humanitarian organizations in places like Haiti and the Philippines to map areas after a natural disaster, using high-resolution embedded cameras, as documented in a recent United Nation report [1]. While research efforts have mostly focused on developing videobased solutions for this task [2], UAV-embedded audio-based localization has received relatively less attention [3], [4], [5], [6], [7]. Though, UAVs equipped with a microphone array could be of critical help to localize people in emergency situations, in particular when video sensors are limited by a lack of visual feedback due to bad lighting conditions (night, fog, etc.)…”

A semblance based TDOA algorithm for sound source localization

“…The problem of estimating the Direction of Arrival (DOA) of a propagating wave plays a fundamental role in many signal processing applications. More recently it has been of great interest for sound source localization, specially in search and rescue scenarios [1].…”

“…For the sound source localization problem, one of the main techniques employed is based on the time difference of arrival (TDOA), i.e., the delay that the propagating wave (sound) arrives at several microphones disposed in different locations [2]- [5]. However, in applications involving Unmanned Aerial Vehicle (UAV) the main problem arises from the ego noise and the fact that the sound source location and the microphones can be in movement [1].…”

Sobre o uso de GCC-PHAT na estimação de DoA para arranjos de dimensão reduzida

“…Os estimadores do segundo tipo apresentam melhor desempenho com arranjos com sensores mais espaçados. Uma das principais vantagens do estimadores de DoAs baseados em TDoA com minimização LSé uma menor complexidade computacional quando comparados aos algoritmos baseados em busca em grade, típicos das técnicas SRP (do inglês, steered response power) [9], favorecendo aplicações em tempo real [10].…”