Bridging CLIP with DINO: Cross-Modal Information Maximization for Online Test-Time Adaptation
cs.CV, cs.LG
Submitted: 2026-11-03
Updated: 2026-11-03
Code: https://github.com/B1ueC1oud/CDIM
Terminology
Sources
- In Search of Lost Domain Generalization
- When Test-Time Adaptation Meets Self-Supervised Models
- Benchmarking Neural Network Robustness to Common Corruptions and Perturbations
- CLIP meets DINO for Tuning Zero-Shot Classifier using Unlabeled Image Collections
- From CLIP to DINO: Visual Encoders Shout in Multi-modal Large Language Models
- Adam: A Method for Stochastic Optimization
- Data or Language Supervision: What Makes CLIP Better than DINO?
- Long-tail learning via logit adjustment
- Towards Stable Test-Time Adaptation in Dynamic Wild World
- DINOv2: Learning Robust Visual Features without Supervision
- Very Deep Convolutional Networks for Large-Scale Image Recognition
- Tent: Fully Test-time Adaptation by Entropy Minimization
- Subspace Alignment for Vision-Language Model Test-time Adaptation
- COME: Test-time adaption by Conservatively Minimizing Entropy
Related papers
- TEA: Text Encoder Alignment for Robust Concept Erasure in Text-to-Image Models
- Do Egocentric Video-Language Models Capture Both Hand- and Object-Centric Cues?
- Aura: Consistent Multi-Subject Video Generation via VLM-Grounded Semantic Alignment
- PVRF: All-in-one Adverse Weather Removal via Prior-modulated and Velocity-constrained Rectified Flow
- DLWM: Diverse Latent World Models for Efficient Multimodal Reasoning
- Out of Sight, Still in Mind: Token Compression for Omni-LLMs