DRONEAUDIONET: Noise Suppression for Drone Audition-based Search and Rescue

2026-08-01

AI Summary

This study addresses the challenge of severe acoustic interference caused by drone rotor noise in aerial auditory tasks such as search and rescue, where signal-to-noise ratios often fall below −10 dB, rendering conventional speech enhancement methods ineffective. To tackle this issue, the authors reformulate a source separation model into a dedicated drone noise estimator and introduce a learnable mask scaling mechanism that extends beyond unit gain, complemented by a residual correction term. This approach represents the first tailored modeling strategy for scenarios dominated by intense drone noise. Experimental results on a public drone auditory dataset demonstrate that the proposed method substantially improves downstream sound classification performance—particularly for human voice detection—and exhibits strong generalization capabilities across unseen drone models and flight configurations.

Abstract

Microphones mounted on UAVs enable aerial acoustic scene analysis applications such as search-and-rescue, wildlife monitoring, and industrial inspection. However, drone rotor noise often dominates the mixture signal at SNRs well below -10 dB, making source recovery extremely challenging. Existing enhancement and source separation methods are typically designed for near-balanced mixtures and degrade substantially in drone audition settings. In this work, we propose DRONEAUDIONET, a drone noise suppression method that reframes a source separation model as a drone noise estimator. To better model drone-dominant mixtures, we introduce a learnable mask-scaling mechanism that allows mask magnitudes beyond unity, together with an additive residual correction term for improved drone estimation and source recovery. We train and evaluate our model on a publicly available drone audition dataset and test generalizability on an out-of-domain dataset with unseen drone hardware and flight modes. Results show that DRONEAUDIONET consistently improves downstream sound classification performance, with the largest gains observed for human vocal sounds. Our findings demonstrate the importance of drone-specific modeling for robust aerial acoustic perception and highlight the potential of source separation methods for real-world drone-assisted search-and-rescue.

Problem

• drone noise
• acoustic scene analysis
• source separation
• noise suppression
• search and rescue

Innovation

• drone noise suppression
• source separation
• mask scaling
• aerial acoustic perception
• search and rescue