Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Variational Few-Shot Learning
Jian Zhang, Chenglong Zhao, Bingbing Ni, Minghao Xu, Xiaokang Yang
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 aggregation of distributions instead of averaging embeddings (ablation) | Variational few-shot · miniImagenet 5-way 5-shot | accuracy (%): 65.62 -> 71.54 | Improved |
| add aggregation of distributions instead of averaging embeddings (ablation) | Variational few-shot · miniImagenet 5-way 1-shot | accuracy (%): 54.89 -> 57.15 | Improved |
| add the learnable-scalar transform function on variance (ablation) | Variational few-shot · miniImagenet 5-way 1-shot | accuracy (%): 53.18 -> 57.15 | Improved |
| add the learnable-scalar transform function on variance (ablation) | Variational few-shot · miniImagenet 5-way 5-shot | accuracy (%): 68.56 -> 71.54 | Improved |
| add two-stage (multi-stage) training with pre-trained features (ablation) | Variational few-shot · miniImagenet 5-way 1-shot | accuracy (%): 55.17 -> 57.15 | Improved |
| add two-stage (multi-stage) training with pre-trained features (ablation) | Variational few-shot · miniImagenet 5-way 5-shot | accuracy (%): 69.13 -> 71.54 | Improved |
| increase target size during training (ablation) | Variational few-shot · miniImagenet 5-way | accuracy (%) | Improved |
| keep training support size equal to test support size (ablation) | Variational few-shot · miniImagenet 5-way | accuracy (%) | Improved |
| swap ConvNet backbone for ResNet-12 | Variational few-shot · miniImagenet 5-way 1-shot | accuracy (%): 57.15 -> 61.23 | Improved |
| swap ConvNet backbone for ResNet-12 | Variational few-shot · miniImagenet 5-way 5-shot | accuracy (%): 71.54 -> 77.69 | Improved |
| add variational distribution learning to MaskNet | MaskNet · Cluttered Omniglot segmentation (4 characters) | IoU (%): 95.8 -> 97.9 | Improved |
| add variational distribution learning to MaskNet | MaskNet · Cluttered Omniglot segmentation (256 characters) | IoU (%): 43.7 -> 43.2 | Hurt |
Closest papers by meaning
- Accelerating Convergence in Bayesian Few-Shot Classification (ICML 2024 · similarity 0.90)
- Bayesian Few-Shot Classification with One-vs-Each Pólya-Gamma Augmented Gaussian Processes (ICLR 2021 · similarity 0.86)
- MetaKernel: Learning Variational Random Features With Limited Labels (TPAMI 2022 · similarity 0.86)
- Provably Improving Generalization of Few-shot models with Synthetic Data (ICML 2025 · similarity 0.85)
- SoftCVI: Contrastive variational inference with self-generated soft labels (ICLR 2025 · similarity 0.85)
- Maxitive Donsker-Varadhan Formulation for Possibilistic Variational Inference (NeurIPS 2026 · similarity 0.85)
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.
