Metraipolis › archive › dc_a6760b45e3d3
Bridging CLIP with DINO: Cross-Modal Information Maximization for Online Test-Time Adaptation — arXiv:2611.04489
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.04489
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-04T21:52:52.842Z
- SHA-256 of the archived text
b5b8a4bcbc5ee558b2b716fda9678eeefb871f7838eedd604456a17672e9d912- Stable URL of this page
https://metraipolis.com/documento/dc_a6760b45e3d3
How to cite
Bridging CLIP with DINO: Cross-Modal Information Maximization for Online Test-Time Adaptation — arXiv:2611.04489 — arXiv (en), https://arxiv.org/abs/2611.04489 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_a6760b45e3d3 (Metraipolis archive dc_a6760b45e3d3, sha256 b5b8a4bcbc5ee558b2b716fda9678eeefb871f7838eedd604456a17672e9d912)
BibTeX
@misc{dca6760b45e3d3,
title = {Bridging CLIP with DINO: Cross-Modal Information Maximization for Online Test-Time Adaptation — arXiv:2611.04489},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.04489},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_a6760b45e3d3; sha256 b5b8a4bcbc5ee558b2b716fda9678eeefb871f7838eedd604456a17672e9d912},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.04489. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Vision-language models such as CLIP enable zero-shot classification from arbitrary class names, but under test-time corruptions, their visual encoder projects distorted features onto incorrect textual anchors with high confidence: a phenomenon we term modality bias. Standard entropy minimization reinforces these overconfident errors, which creates a confirmation-bias loop and collapses the predicted label distribution. We propose an online test-time adaptation (TTA) method that breaks this loop at three levels. At the representation level, a frozen self-supervised backbone (e.g., DINO v2) is combined with CLIP through a cross-covariance logit ensemble. CLIP text prototypes are projected into the feature space of this text-free backbone, so the self-supervised branch classifies native image features and is less affected by the text bias of CLIP. At the optimization level, a gated test-time information-maximization objective pairs per-sample sharpening with batch-level diversity, while an adaptive gate excludes unreliable samples from the gradient. At the prediction level, an online logit adjustment corrects the residual class-distribution skew. CDIM outperforms the compared CLIP-based online TTA baselines on ImageNet-C, CIFAR10-C, and CIFAR100-C, and remains competitive in continual and mixed-domain TTA. Extensive analyses and ablation studies verify the contribution of each of its components. Our code is available at this https URL Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno). Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2611.04489 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-11-04T21:38:49.815Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).