AI Paper Atlas

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

Collaborative Learning With Disentangled Features for Zero-Shot Domain Adaptation

Won Young Jhoo, Jae-Pil Heo

ICCV 2021 · Label-efficient and robust learning › Domain adaptation and generalization · approach: Conditional distribution alignment · 19 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 the attention-based feature refinement stage (ablation)Ours · X-NIST (average over all settings)accuracy (%): 81.72 -> 88.28Improved
add refinement stage (ablation)Ours · X-NIST G->C (ToI MNIST, IrT FashionMNIST)accuracy (%): 68.6 -> 93.3Improved
add refinement stage (ablation)Ours · X-NIST N->G (ToI FashionMNIST, IrT MNIST)accuracy (%): 61.7 -> 83.4Improved
add refinement stage (ablation)Ours · X-NIST G->N (ToI FashionMNIST, IrT MNIST)accuracy (%): 60.6 -> 81.3Improved
add refinement stage (ablation)Ours · X-NIST G->E (ToI MNIST, IrT FashionMNIST)accuracy (%): 84.7 -> 92.9Improved
add refinement stage (ablation)Ours · X-NIST C->G (ToI MNIST, IrT FashionMNIST)accuracy (%): 98.7 -> 98.9No clear effect
add refinement stage (ablation)Ours · X-NIST G->C (ToI NIST, IrT MNIST)accuracy (%): 31.3 -> 45.7Improved
use Ours instead of Wang et al. generative baselineOurs · X-NIST G->C (ToI MNIST, IrT FashionMNIST)accuracy (%): 81.2 -> 93.3Improved
use Ours instead of Source Only ResNet-50Ours · Office-Home Ar->Cl (6 ToI/IrT splits)accuracy (%): 55.1 -> 71.0Improved
use Ours instead of Source Only ResNet-50Ours · Office-Home Rw->Pr (6 ToI/IrT splits)accuracy (%): 83.2 -> 88.5Improved
use Ours instead of Wang et al.Ours · Office-Home Cl->Ar (6 ToI/IrT splits)accuracy (%): 72.6 -> 62.1Hurt

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