Research map / Learning theory and algorithms / Prediction and decision losses
Recovering Top-Two Answers and Confusion Probability in Multi-Choice Crowdsourcing
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 |
|---|---|---|---|
| Train with top-two soft labels instead of hard labels | VGG-19 and ResNet18 · CIFAR10H | test accuracy: 77.64±1.54% -> 79.20±1.04%; 76.49%±1.80% -> 80.58%±2.36% | Helped |
| Train with top-two labels instead of full label distributions | VGG-19 and ResNet18 · CIFAR10H | test accuracy: 78.66±0.97% -> 79.20±1.04%; 80.93%±2.66% -> 80.58%±2.36% | Mixed |
| Train with top-two labels instead of full label distributions under added spammer responses | VGG-19 and ResNet18 · CIFAR10H, spammer ratios 0.1–0.5 | test accuracy: At 0.1, ResNet18: 80.73±0.79% -> 80.18±1.30%; VGG-19: 78.67±1.45% -> 78.90±0.72%. At 0.2, 0.3, 0.4, and 0.5, top-two val | Mixed |
Closest papers by meaning
- No Oops, You Won’t Do It Again: Mechanisms for Self-correction in Crowdsourcing (ICML 2016 · similarity 0.89)
- Optimality of Belief Propagation for Crowdsourced Classification (ICML 2016 · similarity 0.88)
- Label Aggregation for Composite Crowd Tasks by Worker Ability Constraint Satisfaction (AAAI 2025 · similarity 0.86)
- Evaluating LLM-contaminated Crowdsourcing Data Without Ground Truth (NeurIPS 2025 · similarity 0.85)
- Coupled Confusion Correction: Learning from Crowds with Sparse Annotations (AAAI 2024 · similarity 0.85)
- 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.