Deep neural network Based Low-latency Speech Separation with Asymmetric analysis-Synthesis Window Pair
This addresses real-time speech enhancement for applications like assisted hearing, but it is incremental as it builds on existing DNN-based methods with a windowing modification.
The paper tackled low-latency speech separation by proposing an asymmetric analysis-synthesis window pair, which improved separation performance by up to 1.5 dB in SDR while maintaining an 8 ms algorithmic latency.
Time-frequency masking or spectrum prediction computed via short symmetric windows are commonly used in low-latency deep neural network (DNN) based source separation. In this paper, we propose the usage of an asymmetric analysis-synthesis window pair which allows for training with targets with better frequency resolution, while retaining the low-latency during inference suitable for real-time speech enhancement or assisted hearing applications. In order to assess our approach across various model types and datasets, we evaluate it with both speaker-independent deep clustering (DC) model and a speaker-dependent mask inference (MI) model. We report an improvement in separation performance of up to 1.5 dB in terms of source-to-distortion ratio (SDR) while maintaining an algorithmic latency of 8 ms.