Sound Event Localization and Detection task in DCASE 2019

I formulated my Ph.D. work into a research challenge to foster research in the sound event localization and detection domain. As part of the task, we provided open source dataset in both Ambisonic and microphone array formats which contained stationary sound events placed at different spatial locations. Additionally, the state of the art method proposed in my Ph.D. for joined localization, detection, and tracking was provided as the baseline method. I co-organized this task along with Archontis Politis and Tuomas Virtanen.

More details about the task can be found in the official webpage. A total of 22 international teams participated in this task, with 65 Authors from 24 Affiliations (8 Industry). The slides attached below presents some quick highlights of the task.

SELD task highlights, DCASE 2019