Subspace Alignment for Vision-Language Model Test-time Adaptation

• URL: http://arxiv.org/abs/2601.08139v1
• Date: Tue, 13 Jan 2026 02:02:41 GMT
• Title: Subspace Alignment for Vision-Language Model Test-time Adaptation
• Abstract summary: Vision-language models (VLMs) are vulnerable to distribution shifts.<n>Existing test-time adaptation methods rely on zero-shot predictions as pseudo-labels for self-training.<n>We propose SubTTA, which aligns the semantic subspaces of both modalities to enhance zero-shot predictions.
• Score: 82.83192844597593
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Vision-language models (VLMs), despite their extraordinary zero-shot capabilities, are vulnerable to distribution shifts. Test-time adaptation (TTA) emerges as a predominant strategy to adapt VLMs to unlabeled test data on the fly. However, existing TTA methods heavily rely on zero-shot predictions as pseudo-labels for self-training, which can be unreliable under distribution shifts and misguide adaptation due to two fundamental limitations. First (Modality Gap), distribution shifts induce gaps between visual and textual modalities, making cross-modal relations inaccurate. Second (Visual Nuisance), visual embeddings encode rich but task-irrelevant noise that often overwhelms task-specific semantics under distribution shifts. To address these limitations, we propose SubTTA, which aligns the semantic subspaces of both modalities to enhance zero-shot predictions to better guide the TTA process. To bridge the modality gap, SubTTA extracts the principal subspaces of both modalities and aligns the visual manifold to the textual semantic anchor by minimizing their chordal distance. To eliminate visual nuisance, SubTTA projects the aligned visual features onto the task-specific textual subspace, which filters out task-irrelevant noise by constraining visual embeddings within the valid semantic span, and standard TTA is further performed on the purified space to refine the decision boundaries. Extensive experiments on various benchmarks and VLM architectures demonstrate the effectiveness of SubTTA, yielding an average improvement of 2.24% over state-of-the-art TTA methods.

Related papers

• A Multimodal Dataset for Visually Grounded Ambiguity in Machine Translation [53.06004926782424]
  We present VIDA (Visually-Dependent Ambiguity), a dataset of 2,500 carefully curated instances in which resolving an annotated source span requires visual evidence.<n>We also propose Disambiguation-Centric Metrics that use an LLM-as-a-judge classifier to verify whether annotated ambiguous expressions are resolved correctly at the span level.
  arXiv  Detail & Related papers  (2026-05-03T19:55:06Z)
• Gram-Anchored Prompt Learning for Vision-Language Models via Second-Order Statistics [6.327904555091201]
  We argue that relying solely on first-order information is insufficient for robust adaptation.<n>We propose textbfGram-Anchored Prompt Learning (GAPL) for Vision-Language Models via Second-Order Statistics.
  arXiv  Detail & Related papers  (2026-04-05T06:02:07Z)
• Dual-Teacher Distillation with Subnetwork Rectification for Black-Box Domain Adaptation [69.7036779439312]
  Black box domain adaptation represents a highly practical yet extremely challenging setting.<n>We propose a dual teacher distillation with subnetwork rectification (DDSR) model that exploits specific knowledge embedded in black box source models.<n>Our approach demonstrates consistent improvements over state of the art methods, including those using source data or models.
  arXiv  Detail & Related papers  (2026-03-24T07:54:19Z)
• Concept-Guided Fine-Tuning: Steering ViTs away from Spurious Correlations to Improve Robustness [16.541207762213272]
  Vision Transformers (ViTs) often degrade under distribution shifts because they rely on spurious correlations.<n>We introduce a novel finetuning framework that steers model reasoning toward concept-level semantics.<n>Our method improves robustness across multiple ViT-based models.
  arXiv  Detail & Related papers  (2026-03-09T12:31:14Z)
• Aligning the True Semantics: Constrained Decoupling and Distribution Sampling for Cross-Modal Alignment [21.887969353050664]
  Cross-modal alignment is a crucial task in multimodal learning aimed at achieving semantic consistency between vision and language.<n>Traditional algorithms pursue embedding consistency to achieve semantic consistency, ignoring the non-semantic information present in the embedding.<n>We propose a novel cross-modal alignment algorithm via textbfConstrained textbfDecoupling and textbfSampling (CDDS)
  arXiv  Detail & Related papers  (2026-03-05T13:23:49Z)
• Steering Vision-Language-Action Models as Anti-Exploration: A Test-Time Scaling Approach [78.4812458793128]
  We propose textbfTACO, a test-time-scaling framework that applies a lightweight pseudo-count estimator as a high-fidelity verifier of action chunks.<n>Our method resembles the classical anti-exploration principle in offline reinforcement learning (RL), and being gradient-free, it incurs significant computational benefits.
  arXiv  Detail & Related papers  (2025-12-02T14:42:54Z)
• Which Direction to Choose? An Analysis on the Representation Power of Self-Supervised ViTs in Downstream Tasks [43.473390101413166]
  Self-Supervised Learning for Vision Transformers (ViTs) has recently demonstrated considerable potential as a pre-training strategy for a variety of computer vision tasks.<n>This study aims to bridge the gap by systematically evaluating the use of unmodified features across image classification and segmentation tasks.
  arXiv  Detail & Related papers  (2025-09-18T11:46:07Z)
• SUTA-LM: Bridging Test-Time Adaptation and Language Model Rescoring for Robust ASR [58.31068047426522]
  Test-Time Adaptation (TTA) aims to mitigate by adjusting models during inference.<n>Recent work explores combining TTA with external language models, using techniques like beam search rescoring or generative error correction.<n>We propose SUTA-LM, a simple yet effective extension of SUTA, with language model rescoring.<n> Experiments on 18 diverse ASR datasets show that SUTA-LM achieves robust results across a wide range of domains.
  arXiv  Detail & Related papers  (2025-06-10T02:50:20Z)
• Dude: Dual Distribution-Aware Context Prompt Learning For Large Vision-Language Model [27.56988000960972]
  We introduce a new framework based on a dual context of both domain-shared and class-specific contexts. Such dual prompt methods enhance the model's feature representation by joining implicit and explicit factors encoded in Large Language Models. We also formulate the Unbalanced Optimal Transport (UOT) theory to quantify the relationships between constructed prompts and visual tokens.
  arXiv  Detail & Related papers  (2024-07-05T13:15:29Z)
• ViT-Lens: Towards Omni-modal Representations [64.66508684336614]
  ViT-Lens-2 is a framework for representation learning of increasing modalities. We show that ViT-Lens-2 can learn representations for 3D point cloud, depth, audio, tactile and EEG. By seamlessly integrating ViT-Lens-2 into Multimodal Foundation Models, we enable Any-modality to Text and Image Generation.
  arXiv  Detail & Related papers  (2023-11-27T18:52:09Z)
• Boosting Few-shot Fine-grained Recognition with Background Suppression and Foreground Alignment [53.401889855278704]
  Few-shot fine-grained recognition (FS-FGR) aims to recognize novel fine-grained categories with the help of limited available samples. We propose a two-stage background suppression and foreground alignment framework, which is composed of a background activation suppression (BAS) module, a foreground object alignment (FOA) module, and a local to local (L2L) similarity metric. Experiments conducted on multiple popular fine-grained benchmarks demonstrate that our method outperforms the existing state-of-the-art by a large margin.
  arXiv  Detail & Related papers  (2022-10-04T07:54:40Z)
• HSVA: Hierarchical Semantic-Visual Adaptation for Zero-Shot Learning [74.76431541169342]
  Zero-shot learning (ZSL) tackles the unseen class recognition problem, transferring semantic knowledge from seen classes to unseen ones. We propose a novel hierarchical semantic-visual adaptation (HSVA) framework to align semantic and visual domains. Experiments on four benchmark datasets demonstrate HSVA achieves superior performance on both conventional and generalized ZSL.
  arXiv  Detail & Related papers  (2021-09-30T14:27:50Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.