Research map / Label-efficient and robust learning / Semi-supervised learning
BidMatch: Boosting Semi-Supervised Learning by Bi-Dimensional Sample Weight Guidance
Xianling Yang, Zhiwen Yu, Sun Song, Kaixiang 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 CIE to consistency regularization (ablation) | consistency regularization · CIFAR-10 (40 labels) | test error rate (%): 13.08 -> 8.69 | Improved |
| add CIE to consistency regularization (ablation) | consistency regularization · CIFAR-100 (400 labels) | test error rate (%): 51.56 -> 47.97 | Improved |
| add PPR to consistency regularization (ablation) | consistency regularization · CIFAR-10 (40 labels) | test error rate (%): 13.08 -> 4.91 | Improved |
| add PPR to consistency regularization (ablation) | consistency regularization · CIFAR-100 (400 labels) | test error rate (%): 51.56 -> 36.71 | Improved |
| add CIE to PPR (full BidMatch) (ablation) | CR+PPR · CIFAR-10 (40 labels) | test error rate (%): 4.91 -> 4.78 | Improved |
| add CIE to PPR (full BidMatch) (ablation) | CR+PPR · CIFAR-100 (400 labels) | test error rate (%): 36.71 -> 36.27 | Improved |
| use BidMatch instead of FixMatch with very few labels | FixMatch · CIFAR-10 (10 labels) | test error rate (%): 24.79 -> 6.45 | Improved |
| use BidMatch instead of FixMatch | FixMatch · STL-10 (40 labels) | test error rate (%): 35.97 -> 13.75 | Improved |
| use BidMatch instead of SoftMatch | SoftMatch · CIFAR-10 (10 labels) | test error rate (%): 15.77 -> 6.45 | Improved |
| use BidMatch instead of SoftMatch | SoftMatch · CIFAR-100 (10000 labels) | test error rate (%): 22.03 -> 21.24 | Improved |
| use BidMatch instead of SoftMatch | SoftMatch · SVHN (1000 labels) | test error rate (%): 2.01 -> 2.01 | No clear effect |
| use BidMatch instead of FixMatch | FixMatch · ImageNet (100k labels) | top-1 error (%): 43.66 -> 40.04 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.
- FixMatch — FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence (NeurIPS 2020) · better on 11 of 12 shared results
- SoftMatch — SoftMatch: Addressing the Quantity-Quality Tradeoff in Semi-supervised Learning (ICLR 2023) · better on 8 of 9 shared results
- FlexMatch — FlexMatch: Boosting Semi-Supervised Learning with Curriculum Pseudo Labeling (NeurIPS 2021) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- SoftMatch: Addressing the Quantity-Quality Tradeoff in Semi-supervised Learning (ICLR 2023 · similarity 0.90)
- Boosting Semi-Supervised Learning by Exploiting All Unlabeled Data (CVPR 2023 · similarity 0.90)
- CGMatch: A Different Perspective of Semi-supervised Learning (CVPR 2025 · similarity 0.88)
- Bi-Level Optimization for Semi-Supervised Learning with Pseudo-Labeling (AAAI 2025 · similarity 0.88)
- OwMatch: Conditional Self-Labeling with Consistency for Open-World Semi-Supervised Learning (NeurIPS 2024 · similarity 0.87)
- CDMAD: Class-Distribution-Mismatch-Aware Debiasing for Class-Imbalanced Semi-Supervised Learning (CVPR 2024 · 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.
