Research map / Label-efficient and robust learning / Semi-supervised learning
Trash To Treasure: Harvesting OOD Data With Cross-Modal Matching for Open-Set Semi-Supervised Learning
Junkai Huang, Chaowei Fang, Weikai Chen, Zhenhua Chai, Xiaolin Wei, Pengxu Wei, Liang Lin, Guanbin Li
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 rotation-recognition self-supervision using all unlabeled data including OOD (ablation) | Supervised · CIFAR-ID-50 (TIN OOD, 2500 labels) | accuracy (%): 53.18 -> 59.82 | Improved |
| add rotation-recognition self-supervision using all unlabeled data including OOD (ablation) | UDA · CIFAR-ID-50 (TIN OOD, 2500 labels) | accuracy (%): 64.78 -> 70.10 | Improved |
| add rotation-recognition self-supervision to UDA (ablation) | UDA · CIFAR-10 (TIN OOD, 250 labels) | accuracy (%): 88.83 -> 90.69 | Improved |
| add cross-modal matching filtering to UDA+SS (ablation) | UDA+SS · CIFAR-ID-50 (TIN OOD, 2500 labels) | accuracy (%): 70.10 -> 73.19 | Improved |
| add cross-modal matching filtering to UDA+SS (ablation) | UDA+SS · CIFAR-10 (TIN OOD, 250 labels) | accuracy (%): 90.69 -> 91.52 | Improved |
| add entropy minimization on unlabeled data for the matching head (ablation) | UDA+SS+CMF · CIFAR-ID-50 (TIN OOD, 2500 labels) | accuracy (%): 72.38 -> 73.19 | Improved |
| add warm-up stage on all data (ablation) | UDA+SS+CMF · CIFAR-10 (TIN OOD, 250 labels) | accuracy (%): 89.45 -> 91.52 | Improved |
| add warm-up stage on all data (ablation) | UDA+SS+CMF · CIFAR-ID-50 (TIN OOD, 2500 labels) | accuracy (%): 71.81 -> 73.19 | Improved |
| use cross-modal matching instead of confidence-based thresholding (ablation) | UDA+SS · CIFAR-10 (TIN OOD, 250 labels) | accuracy (%): 90.81 -> 91.52 | Improved |
| use cross-modal matching instead of confidence-based thresholding (ablation) | UDA+SS · CIFAR-ID-50 (TIN OOD, 2500 labels) | accuracy (%): 71.16 -> 73.19 | Improved |
| add MTCF to UDA | UDA · CIFAR-ID-50 intra-dataset OOD (2000 labels) | accuracy (%): 66.02 -> 60.06 | Hurt |
| add MTCF to UDA | UDA · CIFAR-ID-50 with TIN OOD (2500 labels) | accuracy (%): 66.65 -> 67.30 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- FixMatch — FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence (NeurIPS 2020) · 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)
- Enhanced OoD Detection through Cross-Modal Alignment of Multi-Modal Representations (CVPR 2025 · similarity 0.87)
- Binary Decomposition: A Problem Transformation Perspective for Open-Set Semi-Supervised Learning (ICML 2024 · similarity 0.86)
- CRoFT: Robust Fine-Tuning with Concurrent Optimization for OOD Generalization and Open-Set OOD Detection (ICML 2024 · similarity 0.86)
- Image Background Serves as Good Proxy for Out-of-distribution Data (ICLR 2024 · similarity 0.86)
- How Does Unlabeled Data Provably Help Out-of-Distribution Detection? (ICLR 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.
