mlsanityData to 7 Oct 2026

Efficient Test-Time Model Adaptation without Forgetting

arXiv 2204.02610 ↗First posted 6 Apr 2022cs.LGcited by 467

Not on the map. Why?

Cited by 30 in Multimodal LLMs / vision-language models, but placed by its own abstract, which didn’t fit.

Abstract (arXiv)

Test-time adaptation (TTA) seeks to tackle potential distribution shifts between training and testing data by adapting a given model w.r.t. any testing sample. This task is particularly important for deep models when the test environment changes frequently. Although some recent attempts have been made to handle this task, we still face two practical challenges: 1) existing methods have to perform backward computation for each test sample, resulting in unbearable prediction cost to many applications; 2) while existing TTA solutions can significantly improve the test performance on out-of-distribution data, they often suffer from severe performance degradation on in-distribution data after TTA (known as catastrophic forgetting). In this paper, we point out that not all the test samples contribute equally to model adaptation, and high-entropy ones may lead to noisy gradients that could disrupt the model. Motivated by this, we propose an active sample selection criterion to identify reliable and non-redundant samples, on which the model is updated to minimize the entropy loss for test-time adaptation. Furthermore, to alleviate the forgetting issue, we introduce a Fisher regularizer to constrain important model parameters from drastic changes, where the Fisher importance is estimated from test samples with generated pseudo labels. Extensive experiments on CIFAR-10-C, ImageNet-C, and ImageNet-R verify the effectiveness of our proposed method.

Key facts. Cited by 467 papers in the map (data through 7 Oct 2026). Builds on 46 of its 59 references that are in the map. First posted on arXiv 6 Apr 2022.

Builds on

46 of 59 references.

References · 46

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2019EfficientNet: Rethinking Model Scaling for Convolutional Neural Networkscited by 6,415
  3. 2016Overcoming catastrophic forgetting in neural networkscited by 4,718
  4. 2019Benchmarking Neural Network Robustness to Common Corruptions and Perturbationscited by 2,857
  5. 2016Learning without Forgettingcited by 2,465
  6. 2017Non-local Neural Networkscited by 2,290
  7. 2020The Many Faces of Robustness: A Critical Analysis of Out-of-Distribution Generalizationcited by 1,842
  8. 2018Unsupervised Representation Learning by Predicting Image Rotationscited by 1,586
  9. 2020WILDS: A Benchmark of in-the-Wild Distribution Shiftscited by 1,366
  10. 2020Energy-based Out-of-distribution Detectioncited by 1,296
  11. 2018Experience Replay for Continual Learningcited by 1,074
  12. 2017StarGAN: Unified Generative Adversarial Networks for Multi-Domain Image-to-Image Translationcited by 1,036
34 more
  1. 2019AugMix: A Simple Data Processing Method to Improve Robustness and Uncertaintycited by 897
  2. 2020Do We Really Need to Access the Source Data? Source Hypothesis Transfer for Unsupervised Domain Adaptationcited by 844
  3. 2020Fast is better than free: Revisiting adversarial trainingcited by 723
  4. 2017Maximum Classifier Discrepancy for Unsupervised Domain Adaptationcited by 682
  5. 2022Continual Test-Time Domain Adaptationcited by 520
  6. 2020Improving robustness against common corruptions by covariate shift adaptationcited by 427
  7. 2021MEMO: Test Time Robustness via Adaptation and Augmentationcited by 376
  8. 2019Orthogonal Gradient Descent for Continual Learningcited by 361
  9. 2019Fast AutoAugmentcited by 243
  10. 2020Evaluating Prediction-Time Batch Normalization for Robustness under Covariate Shiftcited by 235
  11. 2020Universal Source-Free Domain Adaptationcited by 219
  12. 2020Pitfalls of In-Domain Uncertainty Estimation and Ensembling in Deep Learningcited by 208
  13. 2022Improving Out-of-Distribution Robustness via Selective Augmentationcited by 202
  14. 2018Multi-Adversarial Domain Adaptationcited by 195
  15. 2020Inf-Net: Automatic COVID-19 Lung Infection Segmentation from CT Imagescited by 184
  16. 2018Continual Learning of Context-dependent Processing in Neural Networkscited by 178
  17. 2020Dense Regression Network for Video Groundingcited by 176
  18. 2021Source-free Domain Adaptation via Avatar Prototype Generation and Adaptationcited by 111
  19. 2017Medical Image Analysis using Convolutional Neural Networks: A Reviewcited by 99
  20. 2019TEINet: Towards an Efficient Architecture for Video Recognitioncited by 92
  21. 2020RSPNet: Relative Speed Perception for Unsupervised Video Representation Learningcited by 83
  22. 2021Can Subnetwork Structure be the Key to Out-of-Distribution Generalization?cited by 75
  23. 2020On Feature Normalization and Data Augmentationcited by 71
  24. 2021MT3: Meta Test-Time Training for Self-Supervised Test-Time Adaptioncited by 65
  25. 2020Collaborative Unsupervised Domain Adaptation for Medical Image Diagnosiscited by 60
  26. 2021Essentials for Class Incremental Learningcited by 57
  27. 2021SITA: Single Image Test-time Adaptationcited by 57
  28. 2021Confidence-based Out-of-Distribution Detection: A Comparative Study and Analysiscited by 34
  29. 2021Graph Convolutional Module for Temporal Action Localization in Videoscited by 29
  30. 2021Test time Adaptation through Perturbation Robustnesscited by 28
  31. 2021Towards Accurate Text-based Image Captioning with Content Diversity Explorationcited by 25
  32. 2020COVID-DA: Deep Domain Adaptation from Typical Pneumonia to COVID-19cited by 17
  33. 2020Learning to Generate Noise for Multi-Attack Robustnesscited by 17
  34. 2020Disturbance-immune Weight Sharing for Neural Architecture Searchcited by 9

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 11; 2023: 73; 2024: 126; 2025: 118; 2026: 139.

467 papers.

Sort:

By area:

  1. 2023Towards Stable Test-Time Adaptation in Dynamic Wild Worldcited by 409
  2. 2023A Comprehensive Survey on Test-Time Adaptation under Distribution Shiftscited by 353
  3. 2023AdaMerging: Adaptive Model Merging for Multi-Task Learningcited by 248
  4. 2023Robust Test-Time Adaptation in Dynamic Scenarioscited by 185
  5. 2025Dynamic Cheatsheet: Test-Time Learning with Adaptive Memorycited by 146
  6. 2023EcoTTA: Memory-Efficient Continual Test-time Adaptation via Self-distilled Regularizationcited by 123
  7. 2024Entropy is not Enough for Test-Time Adaptation: From the Perspective of Disentangled Factorscited by 106
  8. 2022Test-Time Adaptation via Conjugate Pseudo-labelscited by 98
  9. 2024Test-Time Model Adaptation with Only Forward Passescited by 74
  10. 2023DELTA: degradation-free fully test-time adaptationcited by 72
  11. 2022Back to the Source: Diffusion-Driven Test-Time Adaptationcited by 71
  12. 2023A Comprehensive Survey of Forgetting in Deep Learning Beyond Continual Learningcited by 66
38 more
  1. 2023Universal Test-time Adaptation through Weight Ensembling, Diversity Weighting, and Prior Correctioncited by 65
  2. 2023On Pitfalls of Test-Time Adaptationcited by 63
  3. 2023SoTTA: Robust Test-Time Adaptation on Noisy Data Streamscited by 63
  4. 2023RDumb: A simple approach that questions our progress in continual test-time adaptationcited by 50
  5. 2023AdaNPC: Exploring Non-Parametric Classifier for Test-Time Adaptationcited by 48
  6. 2023Learning Vision-and-Language Navigation from YouTube Videoscited by 47
  7. 2025Test-Time Learning for Large Language Modelscited by 46
  8. 2022A Probabilistic Framework for Lifelong Test-Time Adaptationcited by 44
  9. 2023LaFTer: Label-Free Tuning of Zero-shot Classifier using Language and Unlabeled Image Collectionscited by 43
  10. 2023In Search of Lost Online Test-time Adaptation: A Surveycited by 43
  11. 2023Feature Alignment and Uniformity for Test Time Adaptationcited by 42
  12. 2023Relax Image-Specific Prompt Requirement in SAM: A Single Generic Prompt for Segmenting Camouflaged Objectscited by 40
  13. 2023Towards Open-Set Test-Time Adaptation Utilizing the Wisdom of Crowds in Entropy Minimizationcited by 39
  14. 2022ActMAD: Activation Matching to Align Distributions for Test-Time-Trainingcited by 38
  15. 2023Test-Time Adaptation with CLIP Reward for Zero-Shot Generalization in Vision-Language Modelscited by 37
  16. 2024Frustratingly Easy Test-Time Adaptation of Vision-Language Modelscited by 36
  17. 2023Continual Test-time Domain Adaptation via Dynamic Sample Selectioncited by 35
  18. 2023On the Robustness of Open-World Test-Time Training: Self-Training with Dynamic Prototype Expansioncited by 34
  19. 2023Towards Real-World Test-Time Adaptation: Tri-Net Self-Training with Balanced Normalizationcited by 34
  20. 2022Prototype-Guided Continual Adaptation for Class-Incremental Unsupervised Domain Adaptationcited by 33
  21. 2024DPCore: Dynamic Prompt Coreset for Continual Test-Time Adaptationcited by 33
  22. 2023TTA-COPE: Test-Time Adaptation for Category-Level Object Pose Estimationcited by 32
  23. 2024Just Shift It: Test-Time Prototype Shifting for Zero-Shot Generalization with Vision-Language Modelscited by 31
  24. 2024COME: Test-time adaption by Conservatively Minimizing Entropycited by 31
  25. 2023Continual-MAE: Adaptive Distribution Masked Autoencoders for Continual Test-Time Adaptationcited by 30
  26. 2023Each Test Image Deserves A Specific Prompt: Continual Test-Time Adaptation for 2D Medical Image Segmentationcited by 29
  27. 2023Fast-Slow Test-Time Adaptation for Online Vision-and-Language Navigationcited by 25
  28. 2023TEA: Test-time Energy Adaptationcited by 25
  29. 2024BECoTTA: Input-dependent Online Blending of Experts for Continual Test-time Adaptationcited by 25
  30. 2024Any-Shift Prompting for Generalization over Distributionscited by 24
  31. 2024The Entropy Enigma: Success and Failure of Entropy Minimizationcited by 24
  32. 2026Adaptive Action Chunking at Inference-time for Vision-Language-Action Modelscited by 24
  33. 2023Seasoning Model Soups for Robustness to Adversarial and Natural Distribution Shiftscited by 23
  34. 2025Advances in Multimodal Adaptation and Generalization: From Traditional Approaches to Foundation Modelscited by 23
  35. 2024Protected Test-Time Adaptation via Online Entropy Matching: A Betting Approachcited by 22
  36. 2024TAPT: Test-Time Adversarial Prompt Tuning for Robust Inference in Vision-Language Modelscited by 22
  37. 2022CAFA: Class-Aware Feature Alignment for Test-Time Adaptationcited by 21
  38. 2023Label Shift Adapter for Test-Time Adaptation under Covariate and Label Shiftscited by 21

50 most cited of 467 All 467 (JSON)