Deep Learning Models for Single-Channel Speech Enhancement on Drones

Speech enhancement for drone audition is made challenging by the strong ego-noise from the rotating motors and propellers, which leads to extremely low signal-to-noise ratios (e.g. SNR $< -15$ dB) at onboard microphones. In this paper, we extensively assess the ability of single-channel deep learning approaches to ego-noise reduction on drones. We train twelve representative deep neural network (DNN) models, covering three operation domains (time-frequency magnitude domain, time-frequency complex domain and end-to-end time domain) and three distinct architectures (sequential, encoder-decoder and generative). We critically discuss and compare the performance of these models in extremely low-SNR scenarios, ranging from -30 to 0 dB. We show that time-frequency complex domain and UNet encoder-decoder architectures outperform other approaches on speech enhancement measures while providing a good trade-off with other criteria, such as model size, computation complexity and context length. The best-performing model is a UNet model operating in the time-frequency complex domain, which, at input SNR -15 dB, improves ESTOI from 0.1 to 0.4, PESQ from 1.0 to 1.9 and SI-SDR from -15 dB to 3.7 dB. Based on the insights drawn from these findings, we discuss future research in drone ego-noise reduction.

Automated summary

In this paper, the authors extensively assess the ability of single-channel deep learning approaches to ego-noise reduction on drones, given the challenge posed by strong ego-noise from rotating motors and propellers. The study trains and compares twelve representative deep neural network models, and finds that the time-frequency complex domain and UNet encoder-decoder architectures outperform other approaches in speech enhancement measures while providing a good trade-off with other criteria. The best-performing model is a UNet model operating in the time-frequency complex domain, which significantly improves speech quality at low input signal-to-noise ratios.