The SPCup19 Egonoise Dataset

The first phase of the IEEE Signal Processing Cup 2019 included a bonus task: participants could send their own recordings of drone egonoise to obtain extra points. In robotics, egonoise refers to the noise self-produced by a robot at its acoustic sensors, e.g., by its actuators.

This task was a success, as 10 teams sent very valuable egonoise data. This allowed us to gather a unique database of drone-embedded recordings, which is made publically available here for research and education purpose, as agreed by participating teams.

Please add the following citation to any research paper referring to these data:

Audio-Based Search and Rescue With a Drone: Highlights From the IEEE Signal Processing Cup 2019 Student Competition Antoine Deleforge, Diego Di Carlo, Martin Strauss, Romain Serizel, Lucio Marcenaro, IEEE Signal Processing Magazine 36 (5), 2019. https://hal.archives-ouvertes.fr/hal-02161897/document https://hal.archives-ouvertes.fr/hal-02161897v1/bibtex

Team Idea!_ssu

Drone Model: Quadrotor UAV (Phantom 4 GL300C from DJI)

Number of audio channels: 1

Team Maverick

Drone Model: Quadcopter UAV (YH-19HW from https://amzn.to/2T1bSl3)

Number of audio channels: 4

Team Diagonal_Unloading

Drone Model: Quadrotor UAV (Phantom 4 PRO by DJI)

Number of audio channels: 16

Team ChuMS

Drone Model: Custom built UAV (by Dotterel Technologies)

Number of audio channels: 8

Team LEADS_UAV

Drone Model: Quadrotor UAV (Phantom 3 Advanced by DJI)

Number of audio channels: 1

Team NSS_Chellamma

Drone Model: Self-assembled UAV (see documentation)

Number of audio channels: 1

Team KumamoTech

Drone Model: Quadrotor UAV (enRoute Zion PG560)

Number of audio channels: 16

Team AGH

Drone Model: Quadrotor UAV (unknown)

Number of audio channels: 8

KU Leuven Team 1 & Team 2

Drone Model: Quadrotor UAV (MK EASY Quadro V3)

Number of audio channels: 8

Team Shout COOEE!

Drone Model: Intel Aero Ready-To-Fly Drone

Number of audio channels: 8

Comments are closed.