Research map / Vision-language and multimodal / Human and embodied modeling
Ambiguity Helps: Classification With Disagreements in Crowdsourced Annotations
Viktoriia Sharmanska, Daniel Hernandez-Lobato, Jose Miguel Hernandez-Lobato, Novi Quadrianto
Check 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 |
|---|---|---|---|
| Use annotation-agreement confidence to reduce the influence of disputed training examples | GPC · SUNAttribute (83 scene attributes) | accuracy | Improved |
| Use annotation-agreement confidence in GPC classification | GPC · SUNAttribute (83 scene attributes) | accuracy | Mixed |
| Use annotation-agreement confidence in SVM classification | SVM · SUNAttribute (83 scene attributes) | accuracy | No clear effect |
| Use easy-hard annotation scores as training confidence | GPC and SVM · AwA easy-versus-hard images (excluding seal) | accuracy | Improved |
| Use easy-hard annotation scores as training confidence | GPC and SVM · AwA rat easy vs hard | accuracy: 84.33 ± 1.5 -> 84.00 ± 1.5 | Hurt |
| Use annotation-confidence information in GPCconf rather than image features alone | GPC · AwA leopard easy vs hard | accuracy: 81.67 ± 0.7 -> 82.00 ± 0.7 | Improved |
| Use annotation-confidence information in GPCconf rather than image features alone | GPC · AwA seal easy vs hard | accuracy: 48.00 ± 1.4 -> 48.10 ± 1.2 | No clear effect |
Closest papers by meaning
- Predicting Foreground Object Ambiguity and Efficiently Crowdsourcing the Segmentation(s) (IJCV 2018 · similarity 0.89)
- Seeing Through the Human Reporting Bias: Visual Classifiers From Noisy Human-Centric Labels (CVPR 2016 · similarity 0.87)
- Object Detection Benchmarks are Incomplete: The Role of Label Errors and Annotation Uncertainty (NeurIPS 2026 · similarity 0.84)
- Doomed to Re-Annotate, Forever: The ImageNet Story (NeurIPS 2026 · similarity 0.83)
- Classifier-to-Bias: Toward Unsupervised Automatic Bias Detection for Visual Classifiers (CVPR 2025 · similarity 0.83)
- Coupled Confusion Correction: Learning from Crowds with Sparse Annotations (AAAI 2024 · similarity 0.83)
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.
