AI Paper Atlas

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

Distant Supervised Centroid Shift: A Simple and Efficient Approach to Visual Domain Adaptation

Jian Liang, Ran He, Zhenan Sun, Tieniu Tan

CVPR 2019 · Label-efficient and robust learning › Domain adaptation and generalization · approach: Geometric alignment and synthetic-to-real adaptation · 93 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 centroid-shift subspace adaptation instead of the strongest listed baselineMCS · Office31average accuracy: 76.3 -> 76.5Improved
Use MCS instead of JGSAVGG-FC6 features on Office-Caltech · Office-Caltechaverage accuracy: 83.6 -> 85.8Improved
Use MCS instead of JGSAVGG-FC7 features on Office-Caltech · Office-Caltechaverage accuracy: 82.3 -> 85.2Improved
Use MCS instead of CDAN+EResNet-50 features on Office-Home · Office-Homeaverage accuracy: 65.8 -> 66.3Improved
Apply MCS to RevGrad intermediate featuresRevGrad on Office31 · Office31average accuracy: 81.8 -> 87.8Improved
Apply MCS to DAN intermediate featuresDAN on Office31 · Office31average accuracy: 81.7 -> 87.4Improved
Apply MCS to GTA intermediate featuresGTA on digit recognition · MNIST → USPSaccuracy: 96.5 -> 97.8Improved
Use MCSD instead of StPMulti-source adaptation · PIEaverage accuracy: 54.2 -> 78.0Improved
Use MCSD instead of MsDAMulti-source adaptation · Office31average accuracy: 84.5 -> 86.0Improved
Use MCSD instead of StPMulti-source adaptation · Office-Caltechaverage accuracy: 93.6 -> 94.8Improved
Vary subspace dimensionality (ablation)MCS · Office31 and Office-Home adaptation tasksaccuracyNo clear effect
Increase regularization λ toward very large values (ablation)MCS · Office31 and Office-Home adaptation tasksaccuracyHurt

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