AI Paper Atlas

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

Drop to Adapt: Learning Discriminative Features for Unsupervised Domain Adaptation

Seungmin Lee, Dongwan Kim, Namil Kim, Seong-Gyun Jeong

ICCV 2019 · Label-efficient and robust learning › Domain adaptation and generalization · approach: Adversarial domain adaptation · 191 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 VAT to cDTA (ablation)DTA (ResNet-101) · VisDA-2017 classificationaverage accuracy (%): 77.0 -> 81.2Improved
add VAT to fDTA (ablation)DTA (ResNet-101) · VisDA-2017 classificationaverage accuracy (%): 76.4 -> 80.5Improved
add VAT to fDTA (ablation)DTA (ResNet-50) · VisDA-2017 classificationaverage accuracy (%): 72.5 -> 73.1Improved
add VAT to cDTA (ablation)DTA (ResNet-50) · VisDA-2017 classificationaverage accuracy (%): 74.7 -> 75.4Improved
add fDTA to cDTA + VAT (full DTA) (ablation)DTA (ResNet-101) · VisDA-2017 classificationaverage accuracy (%): 81.2 -> 81.5Improved
add fDTA to cDTA + VAT (full DTA) (ablation)DTA (ResNet-50) · VisDA-2017 classificationaverage accuracy (%): 75.4 -> 76.2Improved
add VAT to cDTA + fDTA (ablation)DTA (ResNet-101) · VisDA-2017 classificationaverage accuracy (%): 77.4 -> 81.5Improved
add fDTA to cDTA (ablation)DTA (ResNet-50) · VisDA-2017 classificationaverage accuracy (%): 74.7 -> 75.0No clear effect
add the full DTA objective to source-only trainingResNet-101 · VisDA-2017 classificationaverage accuracy (%): 50.8 -> 81.5Improved
add DTA to source-only trainingResNet-50 FCN · GTA5 to CityscapesmIoU (%): 24.8 -> 35.8Improved
use DTA instead of ADRResNet-50 FCN · GTA5 to CityscapesmIoU (%): 33.3 -> 35.8Improved
use DTA when STL is the (small) source domainDTA · STL to CIFARtarget accuracy (%): 74.2 -> 72.8Hurt

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