mlsanityData to 7 Oct 2026

Listen, Adapt, Better WER: Source-free Single-utterance Test-time Adaptation for Automatic Speech Recognition

arXiv 2203.14222 ↗First posted 27 Mar 2022eess.AS cs.CL cs.SDcited by 19

Not on the map. Why?

Abstract (arXiv)

Although deep learning-based end-to-end Automatic Speech Recognition (ASR) has shown remarkable performance in recent years, it suffers severe performance regression on test samples drawn from different data distributions. Test-time Adaptation (TTA), previously explored in the computer vision area, aims to adapt the model trained on source domains to yield better predictions for test samples, often out-of-domain, without accessing the source data. Here, we propose the Single-Utterance Test-time Adaptation (SUTA) framework for ASR, which is the first TTA study on ASR to our best knowledge. The single-utterance TTA is a more realistic setting that does not assume test data are sampled from identical distribution and does not delay on-demand inference due to pre-collection for the batch of adaptation data. SUTA consists of unsupervised objectives with an efficient adaptation strategy. Empirical results demonstrate that SUTA effectively improves the performance of the source ASR model evaluated on multiple out-of-domain target corpora and in-domain test samples.

Key facts. Cited by 19 papers in the map (data through 7 Oct 2026). Builds on 24 of its 34 references that are in the map. First posted on arXiv 27 Mar 2022.

Builds on

24 of 34 references.

References · 24

  1. 2020wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representationscited by 3,853
  2. 2019MixMatch: A Holistic Approach to Semi-Supervised Learningcited by 1,390
  3. 2019Common Voice: A Massively-Multilingual Speech Corpuscited by 1,358
  4. 2019SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognitioncited by 1,355
  5. 2020Do We Really Need to Access the Source Data? Source Hypothesis Transfer for Unsupervised Domain Adaptationcited by 844
  6. 2018ADVENT: Adversarial Entropy Minimization for Domain Adaptation in Semantic Segmentationcited by 623
  7. 2021SpeechBrain: A General-Purpose Speech Toolkitcited by 404
  8. 2021W2v-BERT: Combining Contrastive Learning and Masked Language Modeling for Self-Supervised Speech Pre-Trainingcited by 340
  9. 2019Semi-supervised Domain Adaptation via Minimax Entropycited by 333
  10. 2018TED-LIUM 3: twice as much data and corpus repartition for experiments on speaker adaptationcited by 224
  11. 2018A DIRT-T Approach to Unsupervised Domain Adaptationcited by 215
  12. 2019Domain Adaptation for Semantic Segmentation with Maximum Squares Losscited by 138
12 more
  1. 2017AutoDIAL: Automatic DomaIn Alignment Layerscited by 75
  2. 2020Rethinking Evaluation in ASR: Are Our Models Robust Enough?cited by 72
  3. 2019Unsupervised Domain Adaptation using Feature-Whitening and Consensus Losscited by 67
  4. 2021Test-Time Adaptation to Distribution Shift by Confidence Maximization and Input Transformationcited by 67
  5. 2021SITA: Single Image Test-time Adaptationcited by 57
  6. 2021Test-time Batch Statistics Calibration for Covariate Shiftcited by 50
  7. 2021MixNorm: Test-Time Adaptation Through Online Normalization Estimationcited by 47
  8. 2018Domain Adversarial Training for Accented Speech Recognitioncited by 40
  9. 2020Unsupervised Domain Adaptation for Speech Recognition via Uncertainty Driven Self-Trainingcited by 38
  10. 2017Large-Scale Domain Adaptation via Teacher-Student Learningcited by 28
  11. 2021Test time Adaptation through Perturbation Robustnesscited by 28
  12. 2020The RWTH ASR System for TED-LIUM Release 2: Improving Hybrid HMM with SpecAugmentcited by 19

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 3; 2024: 4; 2025: 7; 2026: 5.

19 papers.

Sort:

By area:

  1. 2023A Comprehensive Survey on Test-Time Adaptation under Distribution Shiftscited by 353
  2. 2025E-BATS: Efficient Backpropagation-Free Test-Time Adaptation for Speech Foundation Modelscited by 8
  3. 2023REALM: Robust Entropy Adaptive Loss Minimization for Improved Single-Sample Test-Time Adaptationcited by 7
  4. 2023Test-Time Training for Speechcited by 4
  5. 2024SMILE: Speech Meta In-Context Learning for Low-Resource Language Automatic Speech Recognitioncited by 4
  6. 2024LI-TTA: Language Informed Test-Time Adaptation for Automatic Speech Recognitioncited by 3
  7. 2026Whisper-CD: Accurate Long-Form Speech Recognition using Multi-Negative Contrastive Decodingcited by 3
  8. 2025Recent Trends in Distant Conversational Speech Recognition: A Review of CHiME-7 and 8 DASR Challengescited by 2
  9. 2025Test-Time Training for Speech Enhancementcited by 2
  10. 2025SUTA-LM: Bridging Test-Time Adaptation and Language Model Rescoring for Robust ASRcited by 1
  11. 2025An Investigation of Test-time Adaptation for Audio Classification under Background Noisecited by 1
  12. 2024Self-Train Before You Transcribecited by 0
7 more
  1. 2024Examining Test-Time Adaptation for Personalized Child Speech Recognitioncited by 0
  2. 2025Robust Unsupervised Adaptation of a Speech Recogniser Using Entropy Minimisation and Speaker Codescited by 0
  3. 2025DHAuDS: A Dynamic and Heterogeneous Audio Benchmark for Test-Time Adaptationcited by 0
  4. 2026Test-Time Adaptation for Speech Emotion Recognitioncited by 0
  5. 2026Rethinking Entropy Minimization in Test-Time Adaptation for Autoregressive Modelscited by 0
  6. 2026Robust Multi-Tier Infant-Centered Audio Understanding with Whisper via Structured Speaker Conditioningcited by 0
  7. 2026Test-time adaptation for speech enhancement with an autoregressive speech priorcited by 0