AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Consensus-Driven Active Model Selection

Justin Kay, Grant Van Horn, Subhransu Maji, Daniel Sheldon, Sara Beery

ICCV 2025 · Label-efficient and robust learning › Recognition and long-tail · approach: Deep active learning · 2 citations (OpenAlex)

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 changedWhereMeasuredEffect
use CODA instead of random samplingactive model selection · CoLAcumulative regret at step 100: 368.1 -> 2226.7Hurt
use CODA instead of random samplingactive model selection · CivilCommentscumulative regret at step 100: 140.9 -> 318.6Hurt
use CODA instead of ModelSelectoractive model selection · DomainNet126 painting→realcumulative regret at step 100: 293.4 -> 2.4Improved
use CODA instead of ModelSelectoractive model selection · DomainNet126 sketch→clipartcumulative regret at step 100: 38.1 -> 432.4Hurt
add consensus-informed priors to CODA (ablation)CODA · benchmark taskscumulative regretImproved
add diagonal-weighted prior (vs uniform) to CODA (ablation)CODA · benchmark taskscumulative regretMixed
use expected information gain acquisition instead of random or uncertainty sampling (ablation)CODA · benchmark taskscumulative regretImproved
use uncertainty sampling acquisition inside CODA instead of EIG (ablation)CODA · benchmark taskscumulative regretMixed
use CODA consensus priors only (unsupervised) instead of EnsVunsupervised model selection · 26 tasksregret at step 0Mixed
use CODA instead of EnsVunsupervised model selection · DomainNet126 sketch→realregret at step 0: 4.5 -> 0.1Improved
use CODA instead of EnsVunsupervised model selection · WILDS iwildcamregret at step 0: 0.9 -> 6.1Hurt

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