mlsanityData to 14 Oct 2026

Test-time adaptation for speech enhancement with an autoregressive speech prior

arXiv 2609.03622 ↗First posted 3 Sep 2026cs.SD cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

Test-time adaptation (TTA) offers a promising direction for improving speech enhancement models under mismatched acoustic conditions, without requiring access to labeled target data. In this work, we propose a single-utterance TTA method that regularizes a pretrained speech enhancement model using an autoregressive prior trained on clean speech latent representations extracted from a neural audio codec. Adaptation is performed by minimizing the Kullback-Leibler divergence between the enhanced speech distribution and the clean speech prior. Experiments across multiple noisy speech datasets show consistent improvements in speech quality, particularly under training-testing noise mismatch conditions. Code and audio examples are available online.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 26 of its 32 references that are in the map. First posted on arXiv 3 Sep 2026.

Builds on

26 of 32 references.

References · 26

  1. 2020Conformer: Convolution-augmented Transformer for Speech Recognitioncited by 1,863
  2. 2023High-Fidelity Audio Compression with Improved RVQGANcited by 431
  3. 2021MEMO: Test Time Robustness via Adaptation and Augmentationcited by 377
  4. 2016Professor Forcing: A New Algorithm for Training Recurrent Networkscited by 216
  5. 2022Speech Enhancement and Dereverberation with Diffusion-based Generative Modelscited by 197
  6. 2021DNSMOS P.835: A Non-Intrusive Perceptual Objective Speech Quality Metric to Evaluate Noise Suppressorscited by 172
  7. 2020LibriMix: An Open-Source Dataset for Generalizable Speech Separationcited by 155
  8. 2023ICASSP 2023 Deep Noise Suppression Challengecited by 65
  9. 2024EARS: An Anechoic Fullband Speech Dataset Benchmarked for Speech Enhancement and Dereverberationcited by 64
  10. 2021Source-Free Adaptation to Measurement Shift via Bottom-Up Feature Restorationcited by 50
  11. 2023SELM: Speech Enhancement Using Discrete Tokens and Language Modelscited by 36
  12. 2019A variance modeling framework based on variational autoencoders for speech enhancementcited by 33
14 more
  1. 2022Robustifying Vision Transformer without Retraining from Scratch by Test-Time Class-Conditional Feature Alignmentcited by 27
  2. 2022RemixIT: Continual self-training of speech enhancement models via bootstrapped remixingcited by 20
  3. 2022Listen, Adapt, Better WER: Source-free Single-utterance Test-time Adaptation for Automatic Speech Recognitioncited by 19
  4. 2021Unsupervised Noise Adaptive Speech Enhancement by Discriminator-Constrained Optimal Transportcited by 16
  5. 2018Noise Adaptive Speech Enhancement using Domain Adversarial Trainingcited by 15
  6. 2020On Cross-Corpus Generalization of Deep Learning Based Speech Enhancementcited by 11
  7. 2020NU-GAN: High resolution neural upsampling with GANcited by 11
  8. 2021Test-Time Adaptation Toward Personalized Speech Enhancement: Zero-Shot Learning with Knowledge Distillationcited by 10
  9. 2024Test-time Adaptation for Regression by Subspace Alignmentcited by 10
  10. 2023Assessing the Generalization Gap of Learning-Based Speech Enhancement Systems in Noisy and Reverberant Environmentscited by 6
  11. 2025Modeling strategies for speech enhancement in the latent space of a neural audio codeccited by 6
  12. 2025Test-Time Training for Speech Enhancementcited by 2
  13. 2025Test-Time Adaptation for Speech Enhancement via Domain Invariant Embedding Transformationcited by 1
  14. 2026Test-Time Adaptation For Speech Enhancement Via Mask Polarizationcited by 1

Built on by

Not cited yet.