Convolutional-Recurrent Neural Networks for Speech Enhancement

Han Zhao, Shuayb Zarar, Ivan Tashev, Chin-Hui Lee
2018 2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)  
We propose a novel, end-to-end model based on convolutional and recurrent neural networks for speech enhancement. Our model is purely data-driven: it does not make any assumptions about the type or the stationarity of the noise. In contrast to existing methods that use multilayer perceptrons (MLPs), we employ both convolutional and recurrent neural network architectures. Thus, our approach allows us to exploit local structures in both the spatial and temporal domains. By incorporating prior
more » ... ledge of speech signals into the design of model structures, we build a model that is more data-efficient and achieves better generalization on both seen and unseen noise. Based on experiments with synthetic data, we demonstrate that our model outperforms existing methods, improving PESQ by up to 0.6 on seen noise and 0.64 on unseen noise. Index Termsconvolutional neural networks, recurrent neural networks, speech enhancement, regression model
doi:10.1109/icassp.2018.8462155 dblp:conf/icassp/ZhaoZTL18 fatcat:afhikwhftbbypbkebajlj7mofy