Research map / Label-efficient and robust learning / Semi-supervised learning
DASO: Distribution-Aware Semantics-Oriented Pseudo-Label for Imbalanced Semi-Supervised Learning
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 |
|---|---|---|---|
| Add DASO to FixMatch | FixMatch · CIFAR10-LT, γu=1, N1=500 | accuracy (%): 73.0 ± 3.81 -> 86.6 ± 0.84 | Helped |
| Add DASO to FixMatch | FixMatch · CIFAR10-LT, γu=1/100, N1=500 | accuracy (%): 62.5 ± 0.94 -> 71.0 ± 0.95 | Helped |
| Add DASO to FixMatch | FixMatch · STL10-LT, γl=10, N1=150 | accuracy (%): 56.1 ± 2.32 -> 70.0 ± 1.19 | Helped |
| Add semantic alignment loss (ablation) | DASO · CIFAR10-LT | accuracy: 70.98 -> 75.97 | Helped |
| Add semantic alignment loss (ablation) | DASO · STL10-LT | accuracy: 61.64 -> 70.21 | Helped |
| Use distribution-aware class-adaptive blending rather than constant blending (ablation) | DASO · CIFAR10-LT | accuracy: 72.96 -> 75.97 | Helped |
| Use distribution-aware class-adaptive blending rather than constant blending (ablation) | DASO · STL10-LT | accuracy: 64.21 -> 70.21 | Helped |
| Add semantic pseudo-label contribution (υk=1) instead of relying only on linear pseudo-labels (υk=0) (ablation) | DASO · CIFAR10-LT | accuracy: 73.15; 72.35 | Mixed |
| Add semantic pseudo-label contribution (υk=1) instead of relying only on linear pseudo-labels (υk=0) (ablation) | DASO · STL10-LT | accuracy: 58.51 -> 62.60 | Helped |
| Use balanced prototypes and an EMA encoder (ablation) | DASO · CIFAR10-LT | accuracy: 74.98 -> 75.97 | Helped |
| Use balanced prototypes and an EMA encoder (ablation) | DASO · STL10-LT | accuracy: 68.54 -> 70.21 | Helped |
| Set Tdist to 1.5 rather than 0.3 (ablation) | DASO · CIFAR10-LT | accuracy: 73.97 -> 75.97 | Helped |
Closest papers by meaning
- Distribution Aligning Refinery of Pseudo-label for Imbalanced Semi-supervised Learning (NeurIPS 2020 · similarity 0.92)
- Class-Distribution-Aware Pseudo-Labeling for Semi-Supervised Multi-Label Learning (NeurIPS 2023 · similarity 0.91)
- CDMAD: Class-Distribution-Mismatch-Aware Debiasing for Class-Imbalanced Semi-Supervised Learning (CVPR 2024 · similarity 0.90)
- BaCon: Boosting Imbalanced Semi-supervised Learning via Balanced Feature-Level Contrastive Learning (AAAI 2024 · similarity 0.88)
- Three Heads Are Better than One: Complementary Experts for Long-Tailed Semi-supervised Learning (AAAI 2024 · similarity 0.88)
- Keep It on a Leash: Controllable Pseudo-label Generation Towards Realistic Long-Tailed Semi-Supervised Learning (NeurIPS 2025 · similarity 0.87)
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.