Research map / Label-efficient and robust learning / Semi-supervised learning
Robust Semi-Supervised Learning by Wisely Leveraging Open-Set Data
Yang Yang, N. Jiang, Yi Xu, De‐Chuan Zhan
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 WiseOpen-L (Otsu) | MTC · CIFAR-10 (100 labels) | ID accuracy: 80.51 -> 85.69 | Improved |
| add WiseOpen-E/L variants (mean over variants) | MTC · CIFAR-10 (50 labels) | ID accuracy: +2.98 | Improved |
| add WiseOpen-E/L variants (mean) | OpenMatch · Tiny-ImageNet (50 labels) | ID accuracy: +1.94 | Improved |
| add WiseOpen-E (mean over thresholds) | FixMatch · Tiny-ImageNet (50 labels) | ID accuracy: -0.17 | No clear effect |
| add WiseOpen-E (mean) | FixMatch · CIFAR-100 (100 labels) | ID accuracy: -0.04 | No clear effect |
| add WiseOpen-E (mean) | FreeMatch · CIFAR-10 (400 labels) | ID accuracy: +0.60 | Improved |
| add WiseOpen-E/L (mean) | IOMatch · CIFAR-10 (400 labels) | ID accuracy: +0.00 | No clear effect |
| add WiseOpen (full gradient-variance selection every epoch) | OpenMatch · CIFAR-100 (100 labels) | training time: 10h 59m -> 43h 01m | Hurt |
| use WiseOpen-L instead of full WiseOpen (ablation) | OpenMatch · CIFAR-100 (100 labels) | training time: 43h 01m -> 11h 36m | Improved |
| add WiseOpen (mean over variants) at mismatch ratio 0.4 | OpenMatch · CIFAR-100 (50 labels) | ID accuracy: 70.53 -> 71.88 | Improved |
| add WiseOpen at mismatch ratio 0.7 | OpenMatch · CIFAR-100 (50 labels) | ID accuracy: -0.14 | No clear effect |
| use whole unsupervised loss in gradient-variance selection instead of only L_id or L_ood (ablation) | WiseOpen-E · CIFAR-100 (50 labels) | ID accuracy | 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.
- OpenMatch — OpenMatch: Open-Set Semi-supervised Learning with Open-set Consistency Regularization (NeurIPS 2021) · better on 4 of 5 shared results
- FixMatch — FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence (NeurIPS 2020) · better on 3 of 3 shared results
- FreeMatch — FreeMatch: Self-adaptive Thresholding for Semi-supervised Learning (ICLR 2023) · better on 2 of 2 shared results
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
- Exploration and Exploitation of Unlabeled Data for Open-Set Semi-supervised Learning (IJCV 2024 · similarity 0.91)
- Rethinking Safe Semi-supervised Learning: Transferring the Open-set Problem to A Close-set One (ICCV 2023 · similarity 0.90)
- Let the Void Be Void: Robust Open-Set Semi-Supervised Learning via Selective Non-Alignment (AAAI 2026 · similarity 0.90)
- OpenMatch: Open-Set Semi-supervised Learning with Open-set Consistency Regularization (NeurIPS 2021 · similarity 0.90)
- SCOMatch: Alleviating Overtrusting in Open-set Semi-supervised Learning (ECCV 2024 · similarity 0.90)
- Binary Decomposition: A Problem Transformation Perspective for Open-Set Semi-Supervised Learning (ICML 2024 · similarity 0.86)
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.
