# Deep Noise Suppression Maximizing Non-Differentiable PESQ Mediated by a Non-Intrusive PESQNet

arXiv: https://arxiv.org/abs/2111.03847 · HTML: https://mlsanity.com/paper/2111.03847/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 6 Nov 2021.

## Built on by (most cited first, 7 of 7)

- [The PESQetarian: On the Relevance of Goodhart's Law for Speech Enhancement](https://mlsanity.com/paper/2406.03460/) (2024)
- [Self-Supervised Speech Quality Estimation and Enhancement Using Only Clean Speech](https://mlsanity.com/paper/2402.16321/) (2024)
- [DeepVQE: Real Time Deep Voice Quality Enhancement for Joint Acoustic Echo Cancellation, Noise Suppression and Dereverberation](https://mlsanity.com/paper/2306.03177/) (2023)
- [MetricGAN+/-: Increasing Robustness of Noise Reduction on Unseen Data](https://mlsanity.com/paper/2203.12369/) (2022)
- [Bridging The Multi-Modality Gaps of Audio, Visual and Linguistic for Speech Enhancement](https://mlsanity.com/paper/2501.13375/) (2025)
- [Discriminative Flow Matching: Beyond Time-Conditioning in Generative Restoration via Flow-State Representations](https://mlsanity.com/paper/2609.04525/) (2026)
- [Perceptual Quality Loss or Loss of Perceptual Quality?](https://mlsanity.com/paper/2609.35054/) (2026)

## Abstract (arXiv)

Speech enhancement employing deep neural networks (DNNs) for denoising are called deep noise suppression (DNS). During training, DNS methods are typically trained with mean squared error (MSE) type loss functions, which do not guarantee good perceptual quality. Perceptual evaluation of speech quality (PESQ) is a widely used metric for evaluating speech quality. However, the original PESQ algorithm is non-differentiable, and therefore cannot directly be used as optimization criterion for gradient-based learning. In this work, we propose an end-to-end non-intrusive PESQNet DNN to estimate the PESQ scores of the enhanced speech signal. Thus, by providing a reference-free perceptual loss, it serves as a mediator towards the DNS training, allowing to maximize the PESQ score of the enhanced speech signal. We illustrate the potential of our proposed PESQNet-mediated training on the basis of an already strong baseline DNS. As further novelty, we propose to train the DNS and the PESQNet alternatingly to keep the PESQNet up-to-date and perform well specifically for the DNS under training. Our proposed method is compared to the same DNS trained with MSE-based loss for joint denoising and dereverberation, and the Interspeech 2021 DNS Challenge baseline. Detailed analysis shows that the PESQNet mediation can further increase the DNS performance by about 0.1 PESQ points on synthetic test data and by 0.03 DNSMOS points on real test data, compared to training with the MSE-based loss. Our proposed method also outperforms the Challenge baseline by 0.2 PESQ points on synthetic test data and 0.1 DNSMOS points on real test data.