# Perceptual Quality Loss or Loss of Perceptual Quality?

arXiv: https://arxiv.org/abs/2609.35054 · HTML: https://mlsanity.com/paper/2609.35054/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 18 of its 33 references that are in the map.
- First posted on arXiv 28 Sep 2026.

## Builds on (most cited first, 18 of 18)

- [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://mlsanity.com/paper/2312.00752/) (2023)
- [SDR - half-baked or well done?](https://mlsanity.com/paper/1811.02508/) (2018)
- [Likelihood Training of Schr"odinger Bridge using Forward-Backward SDEs Theory](https://mlsanity.com/paper/2110.11291/) (2021)
- [MetricGAN: Generative Adversarial Networks based Black-box Metric Scores Optimization for Speech Enhancement](https://mlsanity.com/paper/1905.04874/) (2019)
- [Generalization Ability of MOS Prediction Networks](https://mlsanity.com/paper/2110.02635/) (2021)
- [EARS: An Anechoic Fullband Speech Dataset Benchmarked for Speech Enhancement and Dereverberation](https://mlsanity.com/paper/2406.06185/) (2024)
- [An Investigation of Incorporating Mamba for Speech Enhancement](https://mlsanity.com/paper/2405.06573/) (2024)
- [MP-SENet: A Speech Enhancement Model with Parallel Denoising of Magnitude and Phase Spectra](https://mlsanity.com/paper/2305.13686/) (2023)
- [CMGAN: Conformer-based Metric GAN for Speech Enhancement](https://mlsanity.com/paper/2203.15149/) (2022)
- [Schr"odinger Bridge for Generative Speech Enhancement](https://mlsanity.com/paper/2407.16074/) (2024)
- [SCOREQ: Speech Quality Assessment with Contrastive Regression](https://mlsanity.com/paper/2410.06675/) (2024)
- [Investigating Training Objectives for Generative Speech Enhancement](https://mlsanity.com/paper/2409.10753/) (2024)
- [Distillation and Pruning for Scalable Self-Supervised Representation-Based Speech Quality Assessment](https://mlsanity.com/paper/2502.05356/) (2025)
- [The PESQetarian: On the Relevance of Goodhart's Law for Speech Enhancement](https://mlsanity.com/paper/2406.03460/) (2024)
- [Deep Noise Suppression Maximizing Non-Differentiable PESQ Mediated by a Non-Intrusive PESQNet](https://mlsanity.com/paper/2111.03847/) (2021)
- [Too Good to Be True: A Study on Modern Automatic Speech Recognition for the Evaluation of Speech Enhancement](https://mlsanity.com/paper/2605.12107/) (2026)
- [End-to-End Multi-Task Denoising for the Joint Optimization of Perceptual Speech Metrics](https://mlsanity.com/paper/1910.10707/) (2019)
- [Rho-Perfect: Correlation Ceiling For Subjective Evaluation Datasets](https://mlsanity.com/paper/2602.08552/) (2026)

## Abstract (arXiv)

Contemporary deep speech enhancement (SE) models are often trained with specific auxiliary terms in the loss function as a way to improve their performance in terms of perceptual metrics. Nevertheless, a higher score on a perceptual metric does not necessarily correlate with an improved listening experience. Through objective and subjective experiments, we assess the performance of SE models trained with two different types of auxiliary PESQ loss terms. The numerical evaluation on a suite of standard metrics suggests that, while models optimized for PESQ naturally obtain higher PESQ scores in the test set, for most other metrics the scores do not significantly change. In some cases, the PESQ loss even results in worse PESQ scores on mismatched data. A formal listening experiment reveals that the models without a PESQ loss were generally preferred over models that include it, across all settings. Finally, we analyze the relative importance of PESQ in the composite metrics CSIG, CBAK and COVL, and find that PESQ dominates all of them. Our study highlights the perils of over-reliance on PESQ and stresses the importance of a complete evaluation procedure for SE.