aiqbench analysis
This research presents a novel U-Net architecture for acoustic source localization that treats 360-degree audio mapping as a spherical semantic segmentation task rather than discrete angle regression. By leveraging beamformed energy maps and GPS-synchronized drone telemetry, the model achieves improved angular precision while remaining adaptable to various microphone array configurations through transfer learning. This approach advances spatial sound event detection (SELD) beyond traditional methods by utilizing dense spatial representations for robust source identification in complex environments.
Acoustic Imaging for UAV Detection: Dense Beamformed Energy Maps and U-Net SELD
arXiv:2508.00307v4 Announce Type: replace-cross Abstract: We introduce a U-net model for 360{\deg} acoustic source localization formulated as a spherical semantic segmentation task. Rather than regressing discrete direction-of-arrival (DoA) angles, our model segments beamformed audio maps (azimuth & elevation) into regions of active sound presence. Using delay-and-sum (DAS) beamforming on a custom 24-microphone array, we generate signals aligned with drone GPS telemetry to create binary supervision masks. A modified U-Net, trained on frequency-domain representations of these maps, learns to identify spatially distributed source regions while addressing class imbalance via the Tversky loss. Because the network operates on beamformed energy maps, the approach is inherently array-independent and can adapt to different microphone configurations and can be transferred to different microphone configurations with minimal adaptation. The segmentation outputs are post-processed by computing centroids over activated regions, enabling robust DoA estimates. Our dataset includes real-world open-field recordings of a DJI Air 3 drone, synchronized with 360{\deg} video and flight logs across multiple dates and locations. Experimental results show that U-net generalizes across environments, providing improved angular precision, offering a new paradigm for dense spatial audio understanding beyond traditional Sound Source Localization (SSL). We additionally validate the same beamforming-plus-segmentation formulation on the DCASE 2019 TAU Spatial Sound Events benchmark, showing that the approach generalizes beyond drone acoustics to multiclass Sound Event Localization and Detection (SELD) scenarios.
Abstract:We introduce a U-net model for 360° acoustic source localization formulated as a spherical semantic segmentation task. Rather than regressing discrete direction-of-arrival (DoA) angles, our model segments beamformed audio maps (azimuth & elevation) into regions of active sound presence. Using delay-and-sum (DAS) beamforming on a custom 24-microphone array, we generate signals aligned with drone GPS telemetry to create binary supervision masks. A modified U-Net, trained on frequency-domain representations of these maps, learns to identify spatially distributed source regions while addressing class imbalance via the Tversky loss. Because the network operates on beamformed energy maps, the approach is inherently array-independent and can adapt to different microphone configurations and can be transferred to different microphone configurations with minimal adaptation. The segmentation outputs are post-processed by computing centroids over activated regions, enabling robust DoA estimates. Our dataset includes real-world open-field recordings of a DJI Air 3 drone, synchronized with 360° video and flight logs across multiple dates and locations. Experimental results show that U-net generalizes across environments, providing improved angular precision, offering a new paradigm for dense spatial audio understanding beyond traditional Sound Source Localization (SSL). We additionally validate the same beamforming-plus-segmentation formulation on the DCASE 2019 TAU Spatial Sound Events benchmark, showing that the approach generalizes beyond drone acoustics to multiclass Sound Event Localization and Detection (SELD) scenarios.
Audio and Speech Processing (eess.AS); Artificial Intelligence (cs.AI); Sound (cs.SD); Signal Processing (eess.SP)
arXiv:2508.00307 [eess.AS]
(or arXiv:2508.00307v4 [eess.AS] for this version)
Full text under the publisher's open licence · 237 words
Related coverage
News
Quantum Computing Report · Jun 7, 2026
Corporate and Government Early Adopters in Quantum Chemicals & Quantum Life Science
Inside Quantum Technology · Sep 2, 2026
Princeton Team Measures Magnetic Fluctuation Wavelengths below Diffraction Limit
Quantum Zeitgeist · Sep 10, 2026
MalariAI: A Label-Resilient Decoupled Framework for Annotation-Agnostic Cell Segmentation and Explainable Stage Classification in Dense Malaria Blood Smears
arXiv · Aug 5, 2026