mlsanityData to 7 Oct 2026

Continual Test-Time Domain Adaptation

arXiv 2203.13591 ↗First posted 25 Mar 2022cs.CVcited by 520

Not on the map. Why?

Cited by 28 in Multimodal LLMs / vision-language models, but placed by its own abstract, which didn’t fit.

Abstract (arXiv)

Test-time domain adaptation aims to adapt a source pre-trained model to a target domain without using any source data. Existing works mainly consider the case where the target domain is static. However, real-world machine perception systems are running in non-stationary and continually changing environments where the target domain distribution can change over time. Existing methods, which are mostly based on self-training and entropy regularization, can suffer from these non-stationary environments. Due to the distribution shift over time in the target domain, pseudo-labels become unreliable. The noisy pseudo-labels can further lead to error accumulation and catastrophic forgetting. To tackle these issues, we propose a continual test-time adaptation approach~(CoTTA) which comprises two parts. Firstly, we propose to reduce the error accumulation by using weight-averaged and augmentation-averaged predictions which are often more accurate. On the other hand, to avoid catastrophic forgetting, we propose to stochastically restore a small part of the neurons to the source pre-trained weights during each iteration to help preserve source knowledge in the long-term. The proposed method enables the long-term adaptation for all parameters in the network. CoTTA is easy to implement and can be readily incorporated in off-the-shelf pre-trained models. We demonstrate the effectiveness of our approach on four classification tasks and a segmentation task for continual test-time adaptation, on which we outperform existing methods. Our code is available at urlhttps://qin.ee/cotta.

Key facts. Cited by 520 papers in the map (data through 7 Oct 2026). Builds on 50 of its 76 references that are in the map. First posted on arXiv 25 Mar 2022.

Builds on

50 of 76 references.

References · 50

  1. 2016Layer Normalizationcited by 5,579
  2. 2017On Calibration of Modern Neural Networkscited by 4,860
  3. 2016Overcoming catastrophic forgetting in neural networkscited by 4,718
  4. 2016The Cityscapes Dataset for Semantic Urban Scene Understandingcited by 4,427
  5. 2016Wide Residual Networkscited by 3,057
  6. 2021SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformerscited by 2,923
  7. 2019Benchmarking Neural Network Robustness to Common Corruptions and Perturbationscited by 2,857
  8. 2016Aggregated Residual Transformations for Deep Neural Networkscited by 2,540
  9. 2016Learning without Forgettingcited by 2,465
  10. 2016iCaRL: Incremental Classifier and Representation Learningcited by 2,313
  11. 2014Unsupervised Domain Adaptation by Backpropagationcited by 1,906
  12. 2020The Many Faces of Robustness: A Critical Analysis of Out-of-Distribution Generalizationcited by 1,842
38 more
  1. 2017Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real Worldcited by 1,401
  2. 2015Learning Transferable Features with Deep Adaptation Networkscited by 1,323
  3. 2018Continual Lifelong Learning with Neural Networks: A Reviewcited by 1,313
  4. 2017CyCADA: Cycle-Consistent Adversarial Domain Adaptationcited by 1,079
  5. 2019AugMix: A Simple Data Processing Method to Improve Robustness and Uncertaintycited by 897
  6. 2020Do We Really Need to Access the Source Data? Source Hypothesis Transfer for Unsupervised Domain Adaptationcited by 844
  7. 2020RobustBench: a standardized adversarial robustness benchmarkcited by 659
  8. 2013Domain Generalization via Invariant Feature Representationcited by 622
  9. 2018Learning to Adapt Structured Output Space for Semantic Segmentationcited by 584
  10. 2020FDA: Fourier Domain Adaptation for Semantic Segmentationcited by 540
  11. 2020Improving robustness against common corruptions by covariate shift adaptationcited by 427
  12. 2021MEMO: Test Time Robustness via Adaptation and Augmentationcited by 376
  13. 2021DAFormer: Improving Network Architectures and Training Strategies for Domain-Adaptive Semantic Segmentationcited by 299
  14. 2018Training Deep Networks with Synthetic Data: Bridging the Reality Gap by Domain Randomizationcited by 292
  15. 2019A Fourier Perspective on Model Robustness in Computer Visioncited by 288
  16. 2016Revisiting Batch Normalization For Practical Domain Adaptationcited by 285
  17. 2020Universal Source-Free Domain Adaptationcited by 219
  18. 2020Pitfalls of In-Domain Uncertainty Estimation and Ensembling in Deep Learningcited by 208
  19. 2018Progressive Feature Alignment for Unsupervised Domain Adaptationcited by 183
  20. 2021Generalized Source-free Domain Adaptationcited by 175
  21. 2021Source-Free Domain Adaptation for Semantic Segmentationcited by 168
  22. 2020Adversarial Continual Learningcited by 140
  23. 2019Constructing Self-motivated Pyramid Curriculums for Cross-Domain Semantic Segmentation: A Non-Adversarial Approachcited by 103
  24. 2020Test-Time Adaptable Neural Networks for Robust Medical Image Segmentationcited by 102
  25. 2021Domain Impression: A Source Data Free Domain Adaptation Methodcited by 91
  26. 2021Generalize then Adapt: Source-Free Domain Adaptive Semantic Segmentationcited by 81
  27. 2021Domain Adaptive Semantic Segmentation with Self-Supervised Depth Estimationcited by 80
  28. 2020SENTRY: Selective Entropy Optimization via Committee Consistency for Unsupervised Domain Adaptationcited by 78
  29. 2020On Feature Normalization and Data Augmentationcited by 71
  30. 2021Test-Time Adaptation to Distribution Shift by Confidence Maximization and Input Transformationcited by 67
  31. 2020Greedy Policy Search: A Simple Baseline for Learnable Test-Time Augmentationcited by 61
  32. 2020Continual Adaptation of Visual Representations via Domain Randomization and Meta-learningcited by 54
  33. 2017Incremental Adversarial Domain Adaptation for Continually Changing Environmentscited by 52
  34. 2021Test-time Batch Statistics Calibration for Covariate Shiftcited by 50
  35. 2021MixNorm: Test-Time Adaptation Through Online Normalization Estimationcited by 47
  36. 2020Continuous Domain Adaptation with Variational Domain-Agnostic Feature Replaycited by 19
  37. 2021On-target Adaptationcited by 15
  38. 2021Training on Test Data with Bayesian Adaptation for Covariate Shiftcited by 8

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 29; 2023: 105; 2024: 120; 2025: 119; 2026: 147.

520 papers.

Sort:

By area:

  1. 2022Efficient Test-Time Model Adaptation without Forgettingcited by 467
  2. 2023Towards Stable Test-Time Adaptation in Dynamic Wild Worldcited by 409
  3. 2023A Comprehensive Survey on Test-Time Adaptation under Distribution Shiftscited by 353
  4. 2023Robust Test-Time Adaptation in Dynamic Scenarioscited by 185
  5. 2022HRDA: Context-Aware High-Resolution Domain-Adaptive Semantic Segmentationcited by 163
  6. 2022NOTE: Robust Continual Test-time Adaptation Against Temporal Correlationcited by 150
  7. 2023EcoTTA: Memory-Efficient Continual Test-time Adaptation via Self-distilled Regularizationcited by 123
  8. 2022Robust Mean Teacher for Continual and Gradual Test-Time Adaptationcited by 102
  9. 2023TTN: A Domain-Shift Aware Batch Normalization in Test-Time Adaptationcited by 101
  10. 2023Align Your Prompts: Test-Time Prompting with Distribution Alignment for Zero-Shot Generalizationcited by 100
  11. 2022Test-Time Adaptation via Conjugate Pseudo-labelscited by 98
  12. 2022Decorate the Newcomers: Visual Domain Prompt for Continual Test Time Adaptationcited by 98
38 more
  1. 2022Attracting and Dispersing: A Simple Approach for Source-free Domain Adaptationcited by 93
  2. 2022Source-Free Unsupervised Domain Adaptation: A Surveycited by 79
  3. 2024Test-Time Model Adaptation with Only Forward Passescited by 74
  4. 2023DELTA: degradation-free fully test-time adaptationcited by 72
  5. 2023A Comprehensive Survey of Forgetting in Deep Learning Beyond Continual Learningcited by 66
  6. 2023Universal Test-time Adaptation through Weight Ensembling, Diversity Weighting, and Prior Correctioncited by 65
  7. 2024Exponential Moving Average of Weights in Deep Learning: Dynamics and Benefitscited by 65
  8. 2023On Pitfalls of Test-Time Adaptationcited by 63
  9. 2023SoTTA: Robust Test-Time Adaptation on Noisy Data Streamscited by 63
  10. 2023ViDA: Homeostatic Visual Domain Adapter for Continual Test Time Adaptationcited by 58
  11. 2024A Survey on All-in-One Image Restoration: Taxonomy, Evaluation and Future Trendscited by 58
  12. 2022Source-free Depth for Object Pop-outcited by 52
  13. 2023RDumb: A simple approach that questions our progress in continual test-time adaptationcited by 50
  14. 2023C-SFDA: A Curriculum Learning Aided Self-Training Framework for Efficient Source Free Domain Adaptationcited by 49
  15. 2022Revisiting Realistic Test-Time Training: Sequential Inference and Adaptation by Anchored Clusteringcited by 44
  16. 2022A Probabilistic Framework for Lifelong Test-Time Adaptationcited by 44
  17. 2023In Search of Lost Online Test-time Adaptation: A Surveycited by 43
  18. 2023Feature Alignment and Uniformity for Test Time Adaptationcited by 42
  19. 2023Global Features are All You Need for Image Retrieval and Rerankingcited by 41
  20. 2025Centaur: Robust End-to-End Autonomous Driving with Test-Time Trainingcited by 41
  21. 2022Test-Time Robust Personalization for Federated Learningcited by 39
  22. 2023Towards Open-Set Test-Time Adaptation Utilizing the Wisdom of Crowds in Entropy Minimizationcited by 39
  23. 2022ActMAD: Activation Matching to Align Distributions for Test-Time-Trainingcited by 38
  24. 2024Frustratingly Easy Test-Time Adaptation of Vision-Language Modelscited by 36
  25. 2022Video Test-Time Adaptation for Action Recognitioncited by 35
  26. 2023Continual Test-time Domain Adaptation via Dynamic Sample Selectioncited by 35
  27. 2023On the Robustness of Open-World Test-Time Training: Self-Training with Dynamic Prototype Expansioncited by 34
  28. 2023Towards Real-World Test-Time Adaptation: Tri-Net Self-Training with Balanced Normalizationcited by 34
  29. 2024DPCore: Dynamic Prompt Coreset for Continual Test-Time Adaptationcited by 33
  30. 2025R-TPT: Improving Adversarial Robustness of Vision-Language Models through Test-Time Prompt Tuningcited by 33
  31. 2023TTA-COPE: Test-Time Adaptation for Category-Level Object Pose Estimationcited by 32
  32. 2024COME: Test-time adaption by Conservatively Minimizing Entropycited by 31
  33. 2023Prompting Diffusion Representations for Cross-Domain Semantic Segmentationcited by 30
  34. 2023Continual-MAE: Adaptive Distribution Masked Autoencoders for Continual Test-Time Adaptationcited by 30
  35. 2024Unified Language-driven Zero-shot Domain Adaptationcited by 30
  36. 2023Each Test Image Deserves A Specific Prompt: Continual Test-Time Adaptation for 2D Medical Image Segmentationcited by 29
  37. 2023Exploring Sparse Visual Prompt for Domain Adaptive Dense Predictioncited by 28
  38. 2023TeSLA: Test-Time Self-Learning With Automatic Adversarial Augmentationcited by 28

50 most cited of 520 All 520 (JSON)