Two Heads Are Better Than One: A Two-Stage Approach for Monaural Noise Reduction in the Complex Domain
This addresses noise reduction in speech processing, offering a specific improvement for low SNR scenarios, but it is incremental as it builds on existing complex domain methods.
The paper tackled the problem of recovering magnitude and phase information simultaneously in low signal-to-noise ratio conditions for monaural noise reduction, proposing a two-stage algorithm that decouples joint optimization; the results showed significant outperformance over previous baselines in PESQ, ESTOI, and SDR metrics on the WSJ0-SI84 corpus.
In low signal-to-noise ratio conditions, it is difficult to effectively recover the magnitude and phase information simultaneously. To address this problem, this paper proposes a two-stage algorithm to decouple the joint optimization problem w.r.t. magnitude and phase into two sub-tasks. In the first stage, only magnitude is optimized, which incorporates noisy phase to obtain a coarse complex clean speech spectrum estimation. In the second stage, both the magnitude and phase components are refined. The experiments are conducted on the WSJ0-SI84 corpus, and the results show that the proposed approach significantly outperforms previous baselines in terms of PESQ, ESTOI, and SDR.