mlsanityData to 14 Oct 2026

Unleashing the Potential of All Test Samples: Mean-Shift Guided Test-Time Adaptation

arXiv 2507.00462 ↗First posted 1 Jul 2025cs.CVcited by 1

On the map: Multimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP

Abstract (arXiv)

Visual-language models (VLMs) like CLIP exhibit strong generalization but struggle with distribution shifts at test time. Existing training-free test-time adaptation (TTA) methods operate strictly within CLIP's original feature space, relying on high-confidence samples while overlooking the potential of low-confidence ones. We propose MS-TTA, a training-free approach that enhances feature representations beyond CLIP's space using a single-step k-nearest neighbors (kNN) Mean-Shift. By refining all test samples, MS-TTA improves feature compactness and class separability, leading to more stable adaptation. Additionally, a cache of refined embeddings further enhances inference by providing Mean Shift enhanced logits. Extensive evaluations on OOD and cross-dataset benchmarks demonstrate that MS-TTA consistently outperforms state-of-the-art training-free TTA methods, achieving robust adaptation without requiring additional training.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Placed in Prompt learning and adaptation of CLIP, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 46 of its 79 references that are in the map. First posted on arXiv 1 Jul 2025.

Builds on

46 of 79 references.

Closest ancestors · 13

Same direction, newest first

  1. 2025Beyond CLIP Generalization: Against Forward&Backward Forgetting Adapter for Continual Learning of Vision-Language Models12 May
  2. 2025Bayesian Test-Time Adaptation for Vision-Language Models12 Mar
  3. 2024BoostAdapter: Improving Vision-Language Test-Time Adaptation via Regional Bootstrapping20 Oct
  4. 2024Dual Prototype Evolving for Test-Time Generalization of Vision-Language Models16 Oct
  5. 2024DOTA: Distributional Test-Time Adaptation of Vision-Language Models28 Sep
  6. 2024On the test-time zero-shot generalization of vision-language models: Do we really need prompt learning?3 May
  7. 2024Efficient Test-Time Adaptation of Vision-Language Models27 Mar
  8. 2023Diverse Data Augmentation with Diffusions for Effective Test-time Prompt Tuning11 Aug
  9. 2022Test-Time Prompt Tuning for Zero-Shot Generalization in Vision-Language Models15 Sep
  10. 2022Tip-Adapter: Training-free Adaption of CLIP for Few-shot Classification19 Jul
  11. 2022Conditional Prompt Learning for Vision-Language Models10 Mar
  12. 2021Decoupling Zero-Shot Semantic Segmentation15 Dec
1 more
  1. 2021Learning to Prompt for Vision-Language Models2 Sep

Other references · 33

  1. 2017Attention Is All You Needcited by 60,361
  2. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  3. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  4. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 30,186
  5. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  6. 2021Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervisioncited by 3,864
  7. 2013Describing Textures in the Wildcited by 2,125
  8. 2020The Many Faces of Robustness: A Critical Analysis of Out-of-Distribution Generalizationcited by 1,850
  9. 2013Fine-Grained Visual Classification of Aircraftcited by 1,726
  10. 2019Do ImageNet Classifiers Generalize to ImageNet?cited by 1,515
  11. 2017EuroSAT: A Novel Dataset and Deep Learning Benchmark for Land Use and Land Cover Classificationcited by 1,496
  12. 2019Natural Adversarial Examplescited by 1,408
21 more
  1. 2019Learning Robust Global Representations by Penalizing Local Predictive Powercited by 1,013
  2. 2019AugMix: A Simple Data Processing Method to Improve Robustness and Uncertaintycited by 901
  3. 2021MEMO: Test Time Robustness via Adaptation and Augmentationcited by 377
  4. 2022Generalized Category Discoverycited by 211
  5. 2022Parameter-free Online Test-time Adaptationcited by 189
  6. 2023Robust Test-Time Adaptation in Dynamic Scenarioscited by 189
  7. 2022Parametric Classification for Generalized Category Discovery: A Baseline Studycited by 113
  8. 2021Class-agnostic Object Detection with Multi-modal Transformercited by 85
  9. 2021Mean Shift for Self-Supervised Learningcited by 65
  10. 2017Recurrent Pixel Embedding for Instance Groupingcited by 58
  11. 2023Fast-Slow Test-Time Adaptation for Online Vision-and-Language Navigationcited by 25
  12. 2024Historical Test-time Prompt Tuning for Vision Foundation Modelscited by 22
  13. 2023DomainAdaptor: A Novel Approach to Test-time Adaptationcited by 20
  14. 2023PARFormer: Transformer-based Multi-Task Network for Pedestrian Attribute Recognitioncited by 19
  15. 2021MeanShift++: Extremely Fast Mode-Seeking With Applications to Segmentation and Object Trackingcited by 9
  16. 2016Mixture model modal clusteringcited by 8
  17. 2023Convergence Analysis of Mean Shiftcited by 6
  18. 2025Boosting Domain Incremental Learning: Selecting the Optimal Parameters is All You Needcited by 6
  19. 2022GridShift: A Faster Mode-seeking Algorithm for Image Segmentation and Object Trackingcited by 3
  20. 2025Learn by Reasoning: Analogical Weight Generation for Few-Shot Class-Incremental Learningcited by 3
  21. 2025Exploring Audio Cues for Enhanced Test-Time Video Model Adaptationcited by 3

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026ProtoDCS: Towards Robust and Efficient Open-Set Test-Time Adaptation for Vision-Language Modelscited by 1

Similar, not cited

Close in wording, no citation either way

  1. 2026Efficient Test-time Adaptation through Candidate Verification and Divergence Shifts
  2. 2025The Illusion of Progress? A Critical Look at Test-Time Adaptation for Vision-Language Models
  3. 2025ETTA: Efficient Test-Time Adaptation for Vision-Language Models through Dynamic Embedding Updates
  4. 2024CLIPArTT: Adaptation of CLIP to New Domains at Test Time
  5. 2026What Drives Test-Time Adaptation for CLIP? A Controlled Empirical Study from an Update Perspective
  6. 2026On the Limits of Representation-Based Test-Time Adaptation for Multimodal LLMs
  7. 2025Space Rotation with Basis Transformation for Training-free Test-Time Adaptation
  8. 2025CLIPTTA: Robust Contrastive Vision-Language Test-Time Adaptation