AI Paper Atlas

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

CVPR 2016 · Vision-language and multimodal › Human and embodied modeling · approach: Human perception modeling · 23 citations (OpenAlex)

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 changedWhereMeasuredEffect
Use annotation-agreement confidence to reduce the influence of disputed training examplesGPC · SUNAttribute (83 scene attributes)accuracyImproved
Use annotation-agreement confidence in GPC classificationGPC · SUNAttribute (83 scene attributes)accuracyMixed
Use annotation-agreement confidence in SVM classificationSVM · SUNAttribute (83 scene attributes)accuracyNo clear effect
Use easy-hard annotation scores as training confidenceGPC and SVM · AwA easy-versus-hard images (excluding seal)accuracyImproved
Use easy-hard annotation scores as training confidenceGPC and SVM · AwA rat easy vs hardaccuracy: 84.33 ± 1.5 -> 84.00 ± 1.5Hurt
Use annotation-confidence information in GPCconf rather than image features aloneGPC · AwA leopard easy vs hardaccuracy: 81.67 ± 0.7 -> 82.00 ± 0.7Improved
Use annotation-confidence information in GPCconf rather than image features aloneGPC · AwA seal easy vs hardaccuracy: 48.00 ± 1.4 -> 48.10 ± 1.2No clear effect

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy