Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Diversity-Aware Meta Visual Prompting
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use diversity-aware clustering to assign a separate prompt to each homogeneous subset (ablation) | DAM-VP · CUB200, Flowers, CIFAR100, SVHN | Top-1 accuracy: 41.2 -> 63.3; 59.9 -> 81.5; 55.1 -> 66.9; 87.9 -> 88.2 | Helped |
| Initialize cluster prompts with a meta-prompt (ablation) | DAM-VP with diversity-aware partition · CUB200, Flowers, CIFAR100, SVHN | Top-1 accuracy: 63.3 -> 64.6; 81.5 -> 84.1; 66.9 -> 67.2; 88.2 -> 88.4 | Helped |
| Add meta-prompt initialization without diversity-aware partition (ablation) | DAM-VP · CUB200, Flowers, CIFAR100, SVHN | Top-1 accuracy: 41.2 -> 43.9; 59.9 -> 65.7; 55.1 -> 59.4; 87.9 -> 88.1 | Helped |
| Increase the number of meta-training datasets (ablation) | DAM-VP meta-prompt · DTD | accuracy | Helped |
| Vary meta-prompt update step size from 0.1 to 0.7 (ablation) | DAM-VP meta-prompt · DTD | accuracy | No clear effect |
| Increase the sampled subset size used for clustering (ablation) | DAM-VP · Downstream datasets | accuracy | Helped |
| Use a lower clustering threshold to create more prompts (ablation) | DAM-VP · Flowers, head-freezing/missing | accuracy: 64.3 -> 75.7 -> 84.1 -> 88.0 -> 91.1 | Helped |
| Increase clustering prompts by lowering threshold (ablation) | DAM-VP · Flowers, head-freezing/missing | prompt parameters: $0.35$ -> $0.98$ -> $1.82$ -> $3.50$ -> $4.90$ | Hurt |
| Use 10 rather than 50 adaptation epochs (ablation) | DAM-VP · 10-dataset average, head-tuning on ViT-B-22K | accuracy: 85.7 -> 88.5 | Hurt |
| Replace random prompt initialization with meta-prompt initialization (ablation) | DAM-VP · Multiple random seeds | training stability | Helped |
Closest papers by meaning
- LoR-VP: Low-Rank Visual Prompting for Efficient Vision Model Adaptation (ICLR 2025 · similarity 0.90)
- Enhancing Visual Prompting through Expanded Transformation Space and Overfitting Mitigation (NeurIPS 2025 · similarity 0.89)
- AutoVP: An Automated Visual Prompting Framework and Benchmark (ICLR 2024 · similarity 0.89)
- VIPAMIN: Visual Prompt Initialization via Embedding Selection and Subspace Expansion (NeurIPS 2025 · similarity 0.89)
- Query Efficient Black-Box Visual Prompting with Subspace Learning (CVPR 2025 · similarity 0.89)
- Learning to Learn Better Visual Prompts (AAAI 2024 · similarity 0.88)
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.