Research map / Label-efficient and robust learning / Semi-supervised learning
Selective Verification Strategy for Learning From Crowds
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 |
|---|---|---|---|
| use EBM instead of random verification selection (B=25) | semi-supervised DS · Bluebirds | accuracy (%): 91.0 -> 97.2 | Helped |
| use EBM instead of random selection (B=250) | semi-supervised DS · Ages | accuracy (%): 75.6 -> 82.6 | Helped |
| use EBM instead of random selection (B=500) | semi-supervised DS · WebSearch | accuracy (%): 92.0 -> 95.4 | Helped |
| use EBM instead of prediction uncertainty (B=250) | semi-supervised DS · Ages | accuracy (%): 78.2 -> 82.6 | Helped |
| use EBM instead of prediction uncertainty at small B (B=100) | semi-supervised DS · WebSearch | accuracy (%): 89.0 -> 87.4 | No clear effect |
| use EBM instead of uncertainty at B=50 | semi-supervised DS · Ages | accuracy (%): 71.6 -> 71.2 | No clear effect |
| add 25 verified tasks to semi-supervised DS vs unsupervised DS | EBM · Bluebirds | accuracy (%): 88.9 -> 97.2 | Helped |
| add 500 verified tasks vs unsupervised DS | EBM · WebSearch | accuracy (%): 81.5 -> 95.4 | Helped |
| use EBM with EM-based semi-supervised model instead of random | EM-based semi-supervised · Ages | accuracy (%): 75.0 -> 79.1 | Helped |
| use EBM with EM-based model instead of random | EM-based semi-supervised · WebSearch | accuracy (%): 86.4 -> 89.3 | Helped |
| use EBM with EM-based model instead of uncertainty | EM-based semi-supervised · Bluebirds | accuracy (%): 93.5 -> 93.5 | No clear effect |
| use EM-based instead of softmax-based semi-supervised model | EBM · WebSearch | accuracy (%): 95.4 -> 89.3 | Hurt |
Closest papers by meaning
- Mixture of Experts Based Multi-Task Supervise Learning from Crowds (AAAI 2025 · similarity 0.85)
- Coupled Confusion Correction: Learning from Crowds with Sparse Annotations (AAAI 2024 · similarity 0.85)
- Label Aggregation for Composite Crowd Tasks by Worker Ability Constraint Satisfaction (AAAI 2025 · similarity 0.85)
- Efficient Online Crowdsourcing with Complex Annotations (AAAI 2024 · similarity 0.84)
- Unbiased Multi-Label Learning from Crowdsourced Annotations (ICML 2024 · similarity 0.84)
- KFNN: K-Free Nearest Neighbor For Crowdsourcing (NeurIPS 2024 · similarity 0.84)
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.