The official repo of "A Deep Learning Loss Function based on Auditory Power Compression for Speech Enhancement"

This is a loss function for wide-band signals. Although the manuscript was rejected by ICASSP, the method is effective (The main problem appeared in the writing). To ensure the reproducibility of other papers (HGCN), we open-source it.

@misc{wang2021deep,
      title={A Deep Learning Loss Function based on Auditory Power Compression for Speech Enhancement}, 
      author={Tianrui Wang and Weibin Zhu},
      year={2021},
      eprint={2108.11877},
      archivePrefix={arXiv},
      primaryClass={eess.AS}
}