As the cornerstone of other important technologies, such as speech recognition and speech synthesis, speech enhancement is a critical area in audio signal processing. In this paper, a new deep learning structure for speech enhancement is demonstrated. The model introduces a "full" attention mechanism to a bidirectional sequence-to-sequence method to make use of latent information after each focal frame. This is an extension of the previous attention-based RNN method. The proposed bidirectional attention-based architecture achieves better performance in terms of speech quality (PESQ), compared with OM-LSA, CNN-LSTM, T-GSA and the unidirectional attention-based LSTM baseline.
@article{arxiv.2108.12105,
title = {Full Attention Bidirectional Deep Learning Structure for Single Channel Speech Enhancement},
author = {Yuzi Yan and Wei-Qiang Zhang and Michael T. Johnson},
journal= {arXiv preprint arXiv:2108.12105},
year = {2021}
}