Research map / Label-efficient and robust learning / Semi-supervised learning
Pseudo-Loss Confidence Metric for Semi-Supervised Few-Shot Learning
Kai Huang, Jie Geng, Wen Jiang, Xinyang Deng, Zhe Xu
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 selected filter (ablation) | PLCM · step 10 of multi-step training, mini-ImageNet 5-way | accuracy (%): 78.8 -> 83.7 | Improved |
| add selected filter (ablation) | PLCM · step 9 of multi-step training, mini-ImageNet 5-way | accuracy (%): 79.1 -> 83.3 | Improved |
| add selected filter (ablation) | PLCM · step 10 of multi-step training | error rate of selected samples (%): 20.4 -> 13.2 | Improved |
| add selected filter (ablation) | PLCM · step 3 of multi-step training | accuracy (%): 80.5 -> 80.1 | Hurt |
| add class-apart selection to class weight (ablation) | PLCM · mini-ImageNet 5-way 1-shot (base SSFSL) | accuracy (%): 68.28 -> 72.06 | Improved |
| add class weight to class-apart selection (ablation) | PLCM · mini-ImageNet 5-way 1-shot (base SSFSL) | accuracy (%): 71.76 -> 72.06 | Improved |
| add class-apart selection to class weight (ablation) | PLCM · mini-ImageNet 5-way 5-shot (base SSFSL) | accuracy (%): 78.19 -> 83.71 | Improved |
| add class weight to class-apart selection (ablation) | PLCM · mini-ImageNet 5-way 5-shot (base SSFSL) | accuracy (%): 83.03 -> 83.71 | Improved |
| add class-apart selection to class weight (ablation) | ICI · mini-ImageNet 5-way 1-shot (base SSFSL) | accuracy (%): 64.42 -> 70.07 | Improved |
| add class-apart selection to class weight (ablation) | LST · mini-ImageNet 5-way 1-shot (base SSFSL) | accuracy (%): 65.58 -> 70.55 | Improved |
| add multi-step training strategy (ablation) | PLCM · mini-ImageNet 5-way | test accuracy | Improved |
Closest papers by meaning
- Rethinking Confidence Scores and Thresholds in Pseudolabeling-based SSL (ICML 2025 · similarity 0.88)
- An Embarrassingly Simple Approach to Semi-Supervised Few-Shot Learning (NeurIPS 2022 · similarity 0.88)
- Negatives Make a Positive: An Embarrassingly Simple Approach to Semi-Supervised Few-Shot Learning (TPAMI 2023 · similarity 0.87)
- Instance Credibility Inference for Few-Shot Learning (CVPR 2020 · similarity 0.87)
- Keep It on a Leash: Controllable Pseudo-label Generation Towards Realistic Long-Tailed Semi-Supervised Learning (NeurIPS 2025 · similarity 0.86)
- CGMatch: A Different Perspective of Semi-supervised Learning (CVPR 2025 · 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.
