---
title: "Bridging CLIP with DINO: Cross-Modal Information Maximization for Online Test-Time Adaptation"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2611.04489v1"
permalink: "https://feed.astralor.com/items/itm_f984657f971cd38da4c24793"
published: "2026-11-03T08:19:29.000Z"
score: 60
channel: "tech"
tags: ["多模态","Bridging CLIP","DINO","Cross-Modal Information Maximization","Online Test-Time Adaptation","Vision-language","CLIP"]
---

# Bridging CLIP with DINO: Cross-Modal Information Maximization for Online Test-Time Adaptation

## 速览

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 …

## 关键点

- 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 …
- 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.

## 全文

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 https://github.com/B1ueC1oud/CDIM

---

引用前请核对原文;本文由星际流动整理:https://feed.astralor.com/items/itm_f984657f971cd38da4c24793