Vol.63, No.3, 2020, pp.1373-1386, doi:10.32604/cmc.2020.010182
OPEN ACCESS
ARTICLE
Binaural Speech Separation Algorithm Based on Long and Short Time Memory Networks
  • Lin Zhou1, *, Siyuan Lu1, Qiuyue Zhong1, Ying Chen1, 2, Yibin Tang3, Yan Zhou3
1 School of Information Science and Engineering, Southeast University, Nanjing, 210096, China.
2 Department of Psychiatry, Columbia University and NYSPI, New York, 10032, USA.
3 College of Internet of Things Engineering, Hohai University, Changzhou, 213022, China.
* Corresponding Author: Lin Zhou. Email: .
Received 15 February 2020; Accepted 28 February 2020; Issue published 30 April 2020
Abstract
Speaker separation in complex acoustic environment is one of challenging tasks in speech separation. In practice, speakers are very often unmoving or moving slowly in normal communication. In this case, the spatial features among the consecutive speech frames become highly correlated such that it is helpful for speaker separation by providing additional spatial information. To fully exploit this information, we design a separation system on Recurrent Neural Network (RNN) with long short-term memory (LSTM) which effectively learns the temporal dynamics of spatial features. In detail, a LSTM-based speaker separation algorithm is proposed to extract the spatial features in each time-frequency (TF) unit and form the corresponding feature vector. Then, we treat speaker separation as a supervised learning problem, where a modified ideal ratio mask (IRM) is defined as the training function during LSTM learning. Simulations show that the proposed system achieves attractive separation performance in noisy and reverberant environments. Specifically, during the untrained acoustic test with limited priors, e.g., unmatched signal to noise ratio (SNR) and reverberation, the proposed LSTM based algorithm can still outperforms the existing DNN based method in the measures of PESQ and STOI. It indicates our method is more robust in untrained conditions.
Keywords
Binaural speech separation, long and short time memory networks, feature vectors, ideal ratio mask.
Cite This Article
. , "Binaural speech separation algorithm based on long and short time memory networks," Computers, Materials & Continua, vol. 63, no.3, pp. 1373–1386, 2020.
This work is licensed under a Creative Commons Attribution 4.0 International License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.