mlsanityData to 7 Oct 2026

Speech Enhancement using Separable Polling Attention and Global Layer Normalization followed with PReLU

arXiv 2105.02509 ↗First posted 6 May 2021cs.SD cs.AI eess.AScited by 0

Not on the map. Why?

Abstract (arXiv)

Single channel speech enhancement is a challenging task in speech community. Recently, various neural networks based methods have been applied to speech enhancement. Among these models, PHASEN and T-GSA achieve state-of-the-art performances on the publicly opened VoiceBank+DEMAND corpus. Both of the models reach the COVL score of 3.62. PHASEN achieves the highest CSIG score of 4.21 while T-GSA gets the highest PESQ score of 3.06. However, both of these two models are very large. The contradiction between the model performance and the model size is hard to reconcile. In this paper, we introduce three kinds of techniques to shrink the PHASEN model and improve the performance. Firstly, seperable polling attention is proposed to replace the frequency transformation blocks in PHASEN. Secondly, global layer normalization followed with PReLU is used to replace batch normalization followed with ReLU. Finally, BLSTM in PHASEN is replaced with Conv2d operation and the phase stream is simplified. With all these modifications, the size of the PHASEN model is shrunk from 33M parameters to 5M parameters, while the performance on VoiceBank+DEMAND is improved to the CSIG score of 4.30, the PESQ score of 3.07 and the COVL score of 3.73.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 17 of its 30 references that are in the map. First posted on arXiv 6 May 2021.

Builds on

17 of 30 references.

References · 17

  1. 2015Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shiftcited by 6,979
  2. 2016Layer Normalizationcited by 5,579
  3. 2015Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classificationcited by 3,590
  4. 2016Xception: Deep Learning with Depthwise Separable Convolutionscited by 2,320
  5. 2018Conv-TasNet: Surpassing Ideal Time-Frequency Magnitude Masking for Speech Separationcited by 400
  6. 2017SEGAN: Speech Enhancement Generative Adversarial Networkcited by 165
  7. 2019MetricGAN: Generative Adversarial Networks based Black-box Metric Scores Optimization for Speech Enhancementcited by 104
  8. 2019PHASEN: A Phase-and-Harmonics-Aware Speech Enhancement Networkcited by 49
  9. 2017A Wavenet for Speech Denoisingcited by 47
  10. 2019T-GSA: Transformer with Gaussian-weighted self-attention for speech enhancementcited by 34
  11. 2018Speech Denoising with Deep Feature Lossescited by 26
  12. 2018Improved Speech Enhancement with the Wave-U-Netcited by 23
5 more
  1. 2020Improving GANs for Speech Enhancementcited by 22
  2. 2020Exploring the Best Loss Function for DNN-Based Low-latency Speech Enhancement with Temporal Convolutional Networkscited by 10
  3. 2019End-to-End Multi-Task Denoising for joint SDR and PESQ Optimizationcited by 8
  4. 2020Improving Perceptual Quality by Phone-Fortified Perceptual Loss using Wasserstein Distance for Speech Enhancementcited by 3
  5. 2018Multi-Domain Processing via Hybrid Denoising Networks for Speech Enhancementcited by 2

Built on by

Not cited yet.