Research map / Label-efficient and robust learning / Semi-supervised learning
AlphaMatch: Improving Consistency for Semi-Supervised Learning With Alpha-Divergence
Check 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 |
|---|---|---|---|
| replace FixMatch consistency matching with AlphaMatch | SSL with WRN-28-2 · CIFAR-10 (40 labels) | accuracy: 88.71±3.35 -> 91.35±3.38 | Helped |
| replace FixMatch consistency matching with AlphaMatch | SSL with WRN-28-2 · SVHN (40 labels) | accuracy: 92.35±7.65 -> 97.03±0.26 | Helped |
| replace FixMatch consistency matching with AlphaMatch | SSL with WRN-28-2 · CIFAR-100 (400 labels) | accuracy: 59.79±2.94* -> 61.26±3.13* | Helped |
| replace FixMatch consistency matching with AlphaMatch | SSL with WRN-28-2 · STL-10 | accuracy: 89.28±0.63 -> 90.36±0.75 | Helped |
| replace FixMatch consistency matching with AlphaMatch | SSL with DGCNN · ModelNet40 | accuracy: 86.47±0.79 -> 88.32±0.84 | Helped |
| increase augmented examples per image from 1 to 4 (ablation) | AlphaMatch · CIFAR-10 (40 labels) | accuracy: 91.35±3.38 -> 91.63±3.12 | No clear effect |
| increase augmented examples per image from 1 to 4 (ablation) | AlphaMatch · SVHN (40 labels) | accuracy: 97.03±0.26 -> 97.18±0.23 | No clear effect |
| use a larger beta (for example, beta=0.8) instead of a smaller beta (ablation) | AlphaMatch · SVHN (4 labels per class) | accuracy | Hurt |
| increase alpha beyond the best-performing setting (ablation) | AlphaMatch · SVHN (4 labels per class) | accuracy | Hurt |
Closest papers by meaning
- RegMixMatch: Optimizing Mixup Utilization in Semi-Supervised Learning (AAAI 2025 · similarity 0.90)
- Scaling Up Semi-supervised Learning with Unconstrained Unlabelled Data (AAAI 2024 · similarity 0.89)
- OwMatch: Conditional Self-Labeling with Consistency for Open-World Semi-Supervised Learning (NeurIPS 2024 · similarity 0.89)
- CGMatch: A Different Perspective of Semi-supervised Learning (CVPR 2025 · similarity 0.89)
- Erasing the Bias: Fine-Tuning Foundation Models for Semi-Supervised Learning (ICML 2024 · similarity 0.88)
- OTMatch: Improving Semi-Supervised Learning with Optimal Transport (ICML 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.