AI Paper Atlas

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

Robust Optimal Transport with Applications in Generative Modeling and Domain Adaptation

Yogesh Balaji, Rama Chellappa, Soheil Feizi

NeurIPS 2020 · Label-efficient and robust learning › Domain adaptation and generalization · approach: Statistical alignment and hypothesis testing · 35 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 robust OT sample weighting to adversarial alignment (with entropy)adversarial domain adaptation · VISDA-17accuracy (%): 65.5 -> 71.5 (ResNet-50)Improved
add robust OT sample weighting to adversarial alignment (with entropy)adversarial domain adaptation · VISDA-17accuracy (%): 69.3 -> 72.7 (ResNet-101)Improved
add robust OT sample weighting to adversarial alignment (with entropy)adversarial domain adaptation · VISDA-17accuracy (%): 59.5 -> 63.9 (ResNet-18)Improved
add robust OT weighting without entropy regularizeradversarial domain adaptation · VISDA-17accuracy (%): 55.4 -> 62.9 (ResNet-18)Improved
add entropy regularizer on target logits (ablation)robust adversarial adaptation · VISDA-17accuracy (%): 68.6 -> 71.5 (ResNet-50)Improved
use robust WGAN instead of WGAN on data with MNIST/uniform outliersWGAN · CIFAR-10 with outliersFIDImproved
use robust WGAN on clean dataWGAN · CIFAR-10 / CIFAR-100 cleanFID: 21.73 -> 21.57No clear effect
increase rho from 0 to 0.1 (ablation)robust WGAN · CIFAR-10 + MNIST outliersFID: 37.5 -> 29.9Improved
increase rho from 0 to 0.2 (ablation)robust adversarial DA · VISDA-17 ResNet-18accuracy (%): 59.5 -> 63.9Improved

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