Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Bayes-inspired Integration of Pretrained Priors and Few-Shot Evidence for Few-Shot Classification
Mingyang Zhou, Xiaoxuan Zhang, Gang Liu, Yuhong Feng, Xiaoqun Wu, Hao Liao, Rui Mao
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 |
|---|---|---|---|
| add Cache module to MLP+FC (ablation) | BOIF · 11-dataset average, 16-shot, ResNet-50 | accuracy: 80.31 -> 80.61 | Improved |
| add FC module to Cache+MLP (ablation) | BOIF · 11-dataset average, 16-shot, ResNet-50 | accuracy: 79.82 -> 80.61 | Improved |
| add MLP module to Cache+FC (ablation) | BOIF · 11-dataset average, 16-shot, ResNet-50 | accuracy: 79.61 -> 80.61 | Improved |
| add MLP(z^e) only to zero-shot CLIP (ablation) | ZS-CLIP · 11-dataset average, 16-shot, ResNet-50 | accuracy: 58.87 -> 78.67 | Improved |
| add FC(z^e) only to zero-shot CLIP (ablation) | ZS-CLIP · 11-dataset average, 16-shot, ResNet-50 | accuracy: 58.87 -> 78.87 | Improved |
| add Cache module only to zero-shot CLIP (ablation) | ZS-CLIP · 11-dataset average, 16-shot, ResNet-50 | accuracy: 58.87 -> 76.11 | Improved |
| add MLP on top of Cache (ablation) | Cache-only model · 11-dataset average, 16-shot, ResNet-50 | accuracy: 76.11 -> 79.82 | Improved |
| add FC on top of Cache (ablation) | Cache-only model · 11-dataset average, 16-shot, ResNet-50 | accuracy: 76.11 -> 79.61 | Improved |
| add all three modules (Cache, MLP, FC) (ablation) | ZS-CLIP · 11-dataset average, 16-shot, ViT-B/16 | accuracy: 65.52 -> 83.78 | Improved |
| add enriched-feature MLP only (ablation) | ZS-CLIP · 11-dataset average, 16-shot, ViT-B/16 | accuracy: 65.52 -> 82.16 | Improved |
| add Cache module only (ablation) | ZS-CLIP · 11-dataset average, 16-shot, ViT-B/16 | accuracy: 65.52 -> 81.87 | Improved |
| swap to BOIF from LDC (ResNet-50) | LDC · ImageNet-V2 (OOD) | accuracy: 58.03 -> 58.45 | Improved |
Closest papers by meaning
- Bayesian Exploration of Pre-trained Models for Low-shot Image Classification (CVPR 2024 · similarity 0.88)
- Rethinking Few-Shot Adaptation of Vision-Language Models in Two Stages (CVPR 2025 · similarity 0.87)
- Bayesian Test-Time Adaptation for Vision-Language Models (CVPR 2025 · similarity 0.87)
- Pushing the Limit of Fine-Tuning for Few-Shot Learning: Where Feature Reusing Meets Cross-Scale Attention (AAAI 2024 · similarity 0.87)
- Accelerating Convergence in Bayesian Few-Shot Classification (ICML 2024 · similarity 0.86)
- Rethinking Prior Information Generation with CLIP for Few-Shot Segmentation (CVPR 2024 · similarity 0.86)
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.
