Monaural speech enhancement with complex convolutional block attention module and joint time frequency losses

Notes
papers
ai

Graph

Paper

Zhao, Nguyen, and Ma 2021

Notes

ABSTRACT

  • Deep Complex U-Net and CRN achieved SOTA performance for monaural speech enhancement
  • both are encoder decoder structures with skip connections

Bibliography

Zhao, Shengkui, Trung Hieu Nguyen, and Bin Ma. 2021. “Monaural Speech Enhancement with Complex Convolutional Block Attention Module and Joint Time Frequency Losses.” CoRR. http://arxiv.org/abs/2102.01993v1.