Yang Xian
Newcastle University
9 Papers
13 Citations
Yang Xian is an academic researcher from Newcastle University. The author has contributed to research in topics: Speech enhancement & Computer science. The author has an hindex of 3, co-authored 9 publications. Previous affiliations of Yang Xian include Zhengzhou University of Light Industry.
Chat about Author
Papers
Monaural Source Separation in Complex Domain With Long Short-Term Memory Neural Network
TL;DR: The complex signal approximation (cSA), which is operated in the complex domain to utilize the phase information of the desired speech signal to improve the separation performance, is proposed.
Convolutional fusion network for monaural speech enhancement.
TL;DR: In this article, a new convolutional fusion network (CFN) is proposed for monaural speech enhancement by improving model performance, inter-channel dependency, information reuse and parameter efficiency.
18
A Multi-Scale Feature Recalibration Network for End-to-End Single Channel Speech Enhancement
TL;DR: In this article, a multi-scale feature recalibration convolutional encoder-decoder with bidirectional gated recurrent unit (BGRU) architecture was proposed for end-to-end speech enhancement.
Multi-Scale Residual Convolutional Encoder Decoder with Bidirectional Long Short-Term Memory for Single Channel Speech Enhancement
Yang Xian,Yang Sun,Wenwu Wang,Syed Mohsen Naqvi +3 more
- 24 Jan 2021
TL;DR: In this paper, a multi-scale convolutional bidirectional long short-term memory (BLSTM) recurrent neural network was proposed for end-to-end single channel speech enhancement.
13
Estimation of the Number of Sources in Measured Speech Mixtures with Collapsed Gibbs Sampling
Yang Sun,Yang Xian,Pengming Feng,Jonathon A. Chambers,Syed Mohsen Naqvi +4 more
- 01 Dec 2017
TL;DR: Collapsed Gibbs sampling (CGS), a Markov chain Monte Carlo technique, is used to obtain samples from the joint distribution of the speech mixtures, and the Chinese Restaurant Process (CRP) within the framework of the Dirichlet Process (DP) is exploited to cluster samples into different components to finally estimate the number of speakers.
6