AI Paper Atlas

Research map / Label-efficient and robust learning / Domain adaptation and generalization

argmax centroid

Chengyue Gong, Mao Ye, qiang liu

NeurIPS 2021 · Label-efficient and robust learning › Domain adaptation and generalization · approach: Statistical alignment and hypothesis testing · 0 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
add argmax centroid heads on top of SIBSIB · Mini-ImageNet 1-shotaccuracy: 70.0 -> 71.2Improved
add argmax centroid heads on top of SIBSIB · Mini-ImageNet 5-shotaccuracy: 80.0 -> 81.1Improved
add argmax centroid heads on top of IFSLIFSL · Mini-ImageNet 1-shotaccuracy: 73.2±0.3 -> 73.6±0.4Improved
add argmax centroid heads on top of SIBSIB · TieredImageNet 1-shotaccuracy: 81.2±0.6 -> 81.8±0.6Improved
add argmax centroid heads on top of IFSLIFSL · TieredImageNet 5-shotaccuracy: 88.3±0.5 -> 88.5±0.5Improved
select best of n heads instead of averaging ensemble outputsOurs · Mini-ImageNet 1-shot ResNet-18accuracy: 56.7±.3 -> 57.4±.2Improved
select best of n heads vs Bayesian MAMLOurs · Mini-ImageNet 1-shot WRN-28-10accuracy: 57.7±.4 -> 58.1±.4Improved
centroid heads for personalized dialogue vs PAMLOurs · personalized dialoguePPL: 46.35 -> 43.94Improved
centroid heads for personalized dialogue vs PAMLOurs · personalized dialogueBLEU: 0.77 -> 0.91Improved
domain-specific adaptor instead of one shared adaptordomain adaptation baseline · DomainNet subsetaccuracy: 39.5 -> 41.8Improved
argmax centroid adaptors instead of one adaptor per domainOurs · DomainNet subsetaccuracy: 41.8 -> 44.3Improved
random domain-to-model assignment with ensemblingEnsemble · DomainNet subsetaccuracy: 41.8 -> 41.3Hurt

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