mlsanityData to 14 Oct 2026

Hybrid-TTA: Continual Test-time Adaptation via Dynamic Domain Shift Detection

arXiv 2409.08566 ↗First posted 13 Sep 2024cs.CVcited by 5

Not on the map. Why?

Abstract (arXiv)

Continual Test Time Adaptation (CTTA) has emerged as a critical approach for bridging the domain gap between the controlled training environments and the real-world scenarios, enhancing model adaptability and robustness. Existing CTTA methods, typically categorized into Full-Tuning (FT) and Efficient-Tuning (ET), struggle with effectively addressing domain shifts. To overcome these challenges, we propose Hybrid-TTA, a holistic approach that dynamically selects instance-wise tuning method for optimal adaptation. Our approach introduces the Dynamic Domain Shift Detection (DDSD) strategy, which identifies domain shifts by leveraging temporal correlations in input sequences and dynamically switches between FT and ET to adapt to varying domain shifts effectively. Additionally, the Masked Image Modeling based Adaptation (MIMA) framework is integrated to ensure domain-agnostic robustness with minimal computational overhead. Our Hybrid-TTA achieves a notable 1.6%p improvement in mIoU on the Cityscapes-to-ACDC benchmark dataset, surpassing previous state-of-the-art methods and offering a robust solution for real-world continual adaptation challenges.

Key facts. Cited by 5 papers in the map (data through 14 Oct 2026). Builds on 55 of its 73 references that are in the map. First posted on arXiv 13 Sep 2024.

Builds on

55 of 73 references.

References · 55

  1. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 30,186
  2. 2021Swin Transformer: Hierarchical Vision Transformer using Shifted Windowscited by 12,139
  3. 2021Masked Autoencoders Are Scalable Vision Learnerscited by 7,403
  4. 2020Bootstrap your own latent: A new approach to self-supervised Learningcited by 4,750
  5. 2016The Cityscapes Dataset for Semantic Urban Scene Understandingcited by 4,444
  6. 2021SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformerscited by 2,940
  7. 2019Benchmarking Neural Network Robustness to Common Corruptions and Perturbationscited by 2,877
  8. 2020Exploring Simple Siamese Representation Learningcited by 2,866
  9. 2014Unsupervised Domain Adaptation by Backpropagationcited by 1,917
  10. 2021Vision Transformers for Dense Predictioncited by 1,803
  11. 2022Visual Prompt Tuningcited by 1,681
  12. 2021SimMIM: A Simple Framework for Masked Image Modelingcited by 1,153
38 more
  1. 2020Pre-Trained Image Processing Transformercited by 893
  2. 2017A Survey on Multi-Task Learningcited by 840
  3. 2016Playing for Data: Ground Truth from Computer Gamescited by 818
  4. 2022AdaptFormer: Adapting Vision Transformers for Scalable Visual Recognitioncited by 719
  5. 2018End-to-End Multi-Task Learning with Attentioncited by 534
  6. 2022Continual Test-Time Domain Adaptationcited by 525
  7. 2019Pseudo-Labeling and Confirmation Bias in Deep Semi-Supervised Learningcited by 510
  8. 2020Improving robustness against common corruptions by covariate shift adaptationcited by 431
  9. 2019Contrastive Adaptation Network for Unsupervised Domain Adaptationcited by 407
  10. 2021DAFormer: Improving Network Architectures and Training Strategies for Domain-Adaptive Semantic Segmentationcited by 300
  11. 2020Universal Source-Free Domain Adaptationcited by 220
  12. 2022Test-Time Training with Masked Autoencoderscited by 212
  13. 2022MIC: Masked Image Consistency for Context-Enhanced Domain Adaptationcited by 178
  14. 2021Generalized Source-free Domain Adaptationcited by 175
  15. 2022HRDA: Context-Aware High-Resolution Domain-Adaptive Semantic Segmentationcited by 164
  16. 2022NOTE: Robust Continual Test-time Adaptation Against Temporal Correlationcited by 151
  17. 2023EcoTTA: Memory-Efficient Continual Test-time Adaptation via Self-distilled Regularizationcited by 123
  18. 2023TTN: A Domain-Shift Aware Batch Normalization in Test-Time Adaptationcited by 101
  19. 2022Decorate the Newcomers: Visual Domain Prompt for Continual Test Time Adaptationcited by 98
  20. 2022Visual Prompt Tuning for Test-time Domain Adaptationcited by 97
  21. 2023PromptStyler: Prompt-driven Style Generation for Source-free Domain Generalizationcited by 59
  22. 2020Rain rendering for evaluating and improving robustness to bad weathercited by 57
  23. 2022A Probabilistic Framework for Lifelong Test-Time Adaptationcited by 46
  24. 2022PiPa: Pixel- and Patch-wise Self-supervised Learning for Domain Adaptative Semantic Segmentationcited by 41
  25. 2024Domain-Agnostic Mutual Prompting for Unsupervised Domain Adaptationcited by 32
  26. 2023Continual-MAE: Adaptive Distribution Masked Autoencoders for Continual Test-Time Adaptationcited by 30
  27. 2022Domain Generalization by Learning and Removing Domain-specific Featurescited by 28
  28. 2023Exploring Sparse Visual Prompt for Domain Adaptive Dense Predictioncited by 28
  29. 2020Adaptive confidence thresholding for monocular depth estimationcited by 25
  30. 2022On the Road to Online Adaptation for Semantic Image Segmentationcited by 23
  31. 2024Rethinking Multi-domain Generalization with A General Learning Objectivecited by 23
  32. 2023Label Shift Adapter for Test-Time Adaptation under Covariate and Label Shiftscited by 21
  33. 2022Online Domain Adaptation for Semantic Segmentation in Ever-Changing Conditionscited by 16
  34. 2023Improving Domain Generalization with Domain Relationscited by 14
  35. 2023To Adapt or Not to Adapt? Real-Time Adaptation for Semantic Segmentationcited by 14
  36. 2024Salience-Based Adaptive Masking: Revisiting Token Dynamics for Enhanced Pre-trainingcited by 14
  37. 2024Emerging Property of Masked Token for Effective Pre-trainingcited by 10
  38. 2023Complementary Domain Adaptation and Generalization for Unsupervised Continual Domain Shift Learningcited by 9

50 most cited of 55 All 55 (JSON)

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 1; 2026: 4.

5 papers.

  1. 2026Dual-level Adaptation for Multi-Object Tracking: Building Test-Time Calibration from Experience and Intuitioncited by 2
  2. 2025RobIA: Robust Instance-aware Continual Test-time Adaptation for Deep Stereocited by 1
  3. 2026Continual Test-Time Adaptation in Computer Vision: Methods, Benchmarks, and Future Directionscited by 1
  4. 2026Remembering by Reconstructing: Domain Incremental Learning With Test-Time Training on Video Streamscited by 0
  5. 2026When to Adapt: Multi-Signal Domain Shift Detection for Efficient Training-Free Adaptation in Open-Vocabulary Segmentationcited by 0