Documents
Poster
Speech enhancement with neural homomorphic synthesis
![](https://sigport.org/sites/default/files/styles/home/public/WX20220507-120258%402x.png?itok=XnE9pHH-)
- Citation Author(s):
- Submitted by:
- wenbin jiang
- Last updated:
- 7 May 2022 - 12:04am
- Document Type:
- Poster
- Document Year:
- 2022
- Event:
- Presenters:
- Wenbin Jiang
- Paper Code:
- AUD-11.3
- Categories:
- Log in to post comments
Most deep learning-based speech enhancement methods operate directly on time-frequency representations or learned features without making use of the model of speech production. This work proposes a new speech enhancement method based on neural homomorphic synthesis. The speech signal is firstly decomposed into excitation and vocal tract with complex cepstrum analysis. Then, two complex-valued neural networks are applied to estimate the target complex spectrum of the decomposed components. Finally, the time-domain speech signal is synthesized from the estimated excitation and vocal tract. Furthermore, we investigated numerous loss functions and found that the multi-resolution STFT loss, commonly used in the TTS vocoder, benefits speech enhancement. Experimental results demonstrate that the proposed method outperforms existing state-of-the-art complex-valued neural network-based methods in terms of both PESQ and eSTOI.