Research map / Vision-language and multimodal / CLIP and zero-shot
Density-Aware Translation of Spurious Correlations in Zero-Shot VLMs
Afsaneh Hasanebrahimi, Hanxun Huang, Christopher Leckie, Sarah Monazam Erfani
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| apply DAT density-aware translation | CLIP ViT-B/32 · Waterbirds | worst-group accuracy: 41.37 -> 75.08 | Improved |
| apply DAT density-aware translation | CLIP ViT-L/14 · Waterbirds | worst-group accuracy: 31.93 -> 83.33 | Improved |
| apply DAT density-aware translation | CLIP ResNet50 · Waterbirds | worst-group accuracy: 35.36 -> 75.08 | Improved |
| apply DAT density-aware translation | CLIP ViT-B/32 · CelebA | worst-group accuracy: 78.89 -> 78.53 | No clear effect |
| apply DAT density-aware translation | CLIP ViT-L/14 · CelebA | worst-group accuracy: 73.35 -> 84.94 | Improved |
| apply DAT density-aware translation | BiomedCLIP · COVID-19 | worst-group accuracy: 44.83 -> 65.22 | Improved |
| apply DAT density-aware translation | CLIP ViT-L/14 · FMoW | worst-group accuracy: 18.06 -> 27.75 | Improved |
| apply DAT density-aware translation | ALIGN · Waterbirds | worst-group accuracy: 50.3 -> 73.36 | Improved |
| apply DAT density-aware translation | AltCLIP · Waterbirds | worst-group accuracy: 35.8 -> 81.31 | Improved |
| use SLOF density instead of DAO (MLE, k=30) (ablation) | DAT · Waterbirds ViT-B/32 | worst-group accuracy: 71.88 -> 75.08 | Improved |
| use SLOF density instead of k-NN distance (k=10) (ablation) | DAT · Waterbirds ViT-B/32 | worst-group accuracy: 71.49 -> 75.08 | Improved |
| use DAT instead of TIE for compute | CLIP ViT-B/32 · CelebA | computation time (s): 970 -> 203 | Improved |
Closest papers by meaning
- Interpreting and Analysing CLIP's Zero-Shot Image Classification via Mutual Knowledge (NeurIPS 2024 · similarity 0.88)
- Building Vision-Language Models on Solid Foundations with Masked Distillation (CVPR 2024 · similarity 0.88)
- Unlocking Fine-Grained Perception in CLIP via Structurally-Aware Latent Masked Modeling (NeurIPS 2026 · similarity 0.88)
- Mining Fine-Grained Image-Text Alignment for Zero-Shot Captioning via Text-Only Training (AAAI 2024 · similarity 0.88)
- DH-Set: Improving Vision-Language Alignment with Diverse and Hybrid Set-Embeddings Learning (CVPR 2025 · similarity 0.87)
- DINOv2 Meets Text: A Unified Framework for Image- and Pixel-Level Vision-Language Alignment (CVPR 2025 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
