Research map / Label-efficient and robust learning / Semi-supervised learning
OpenMatch: Open-Set Semi-supervised Learning with Open-set Consistency Regularization
Read the paperCheck 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 soft open-set consistency regularization (SOCR) to the OVA outlier detector (ablation) | OpenMatch without FixMatch · CIFAR10 50 labels | AUROC: 60.5 -> 81.3 | Helped |
| add SOCR (ablation) | OpenMatch without FixMatch · CIFAR10 400 labels | AUROC: 75.8 -> 96.8 | Helped |
| add SOCR (ablation) | OpenMatch without FixMatch · CIFAR100 80/20 50 labels | AUROC: 70.4 -> 78.9 | Helped |
| apply soft consistency to outlier detector rather than closed-set classifier (ablation) | OpenMatch · CIFAR10 50 labels | AUROC: 70.6 -> 81.3 | Helped |
| apply soft consistency to closed-set classifier versus none (ablation) | OpenMatch · CIFAR10 50 labels | AUROC: 60.5 -> 70.6 | Helped |
| add FixMatch on pseudo-inliers only (ablation) | OpenMatch with SOCR · CIFAR10 | outlier separation | Helped |
| apply FixMatch to all unlabeled data (no outlier filtering) | FixMatch · CIFAR10 50 labels | error rate: 35.7 -> 43.2 | Hurt |
| SimCLR-pretrained initialization instead of random (ablation) | OpenMatch · CIFAR10 50 labels | error rate / AUROC: 10.4 -> 10.6 | No clear effect |
| add SimCLR loss on labeled or unlabeled data (ablation) | Labeled-only outlier detector · CIFAR10 50 labels | AUROC: 63.9 -> 62.1 | Hurt |
| add rotation prediction on labeled data (ablation) | Labeled-only outlier detector · CIFAR10 50 labels | AUROC: 63.9 -> 70.0 | Helped |
| add rotation prediction on labeled data (ablation) | Labeled-only outlier detector · CIFAR100 50 labels | AUROC: 70.3 -> 67.8 | Hurt |
| add rotation prediction on unlabeled data (ablation) | Labeled-only outlier detector · CIFAR10 50 labels | AUROC: 63.9 -> 64.0 | No clear effect |
Closest papers by meaning
- Binary Decomposition: A Problem Transformation Perspective for Open-Set Semi-Supervised Learning (ICML 2024 · similarity 0.90)
- Unknown-Aware Graph Regularization for Robust Semi-supervised Learning from Uncurated Data (AAAI 2024 · similarity 0.87)
- OwMatch: Conditional Self-Labeling with Consistency for Open-World Semi-Supervised Learning (NeurIPS 2024 · similarity 0.87)
- ANEDL: Adaptive Negative Evidential Deep Learning for Open-Set Semi-supervised Learning (AAAI 2024 · similarity 0.86)
- Scaling Up Semi-supervised Learning with Unconstrained Unlabelled Data (AAAI 2024 · similarity 0.86)
- How Does Unlabeled Data Provably Help Out-of-Distribution Detection? (ICLR 2024 · 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.