Self-attending RNN for Speech Enhancement to Improve Cross-corpus Generalization
arXiv:2105.12831 · doi:10.1109/TASLP.2022.3161143
Abstract
Deep neural networks (DNNs) represent the mainstream methodology for supervised speech enhancement, primarily due to their capability to model complex functions using hierarchical representations. However, a recent study revealed that DNNs trained on a single corpus fail to generalize to untrained corpora, especially in low signal-to-noise ratio (SNR) conditions. Developing a noise, speaker, and corpus independent speech enhancement algorithm is essential for real-world applications. In this study, we propose a self-attending recurrent neural network, or attentive recurrent network (ARN), for time-domain speech enhancement to improve cross-corpus generalization. ARN comprises of recurrent neural networks (RNNs) augmented with self-attention blocks and feedforward blocks. We evaluate ARN on different corpora with nonstationary noises in low SNR conditions. Experimental results demonstrate that ARN substantially outperforms competitive approaches to time-domain speech enhancement, such as RNNs and dual-path ARNs. Additionally, we report an important finding that the two popular approaches to speech enhancement: complex spectral mapping and time-domain enhancement, obtain similar results for RNN and ARN with large-scale training. We also provide a challenging subset of the test set used in this study for evaluating future algorithms and facilitating direct comparisons.
Accepted for publication in IEEE/ACM Transactions on Audio, Speech and Language Processing
References in corpus (5)
- Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
- MetricGAN: Generative Adversarial Networks based Black-box Metric Scores Optimization for Speech Enhancement
- Single Headed Attention RNN: Stop Thinking With Your Head
- Dual-path Self-Attention RNN for Real-Time Speech Enhancement
- Complex Spectral Mapping With Attention Based Convolution Recurrent Neural Network for Speech Enhancement