Sound Event Detection and Localization with Distance Estimation
This work addresses the limitation of incomplete sound source information in SELD for applications like audio scene analysis, but it is incremental as it builds on existing methods by adding distance estimation.
The paper tackled the problem of extending Sound Event Detection and Localization (SELD) to include distance estimation for full 3D sound source positioning, and the result showed that 3D SELD could be performed without degrading performance in detection and direction-of-arrival estimation.
Sound Event Detection and Localization (SELD) is a combined task of identifying sound events and their corresponding direction-of-arrival (DOA). While this task has numerous applications and has been extensively researched in recent years, it fails to provide full information about the sound source position. In this paper, we overcome this problem by extending the task to Sound Event Detection, Localization with Distance Estimation (3D SELD). We study two ways of integrating distance estimation within the SELD core - a multi-task approach, in which the problem is tackled by a separate model output, and a single-task approach obtained by extending the multi-ACCDOA method to include distance information. We investigate both methods for the Ambisonic and binaural versions of STARSS23: Sony-TAU Realistic Spatial Soundscapes 2023. Moreover, our study involves experiments on the loss function related to the distance estimation part. Our results show that it is possible to perform 3D SELD without any degradation of performance in sound event detection and DOA estimation.