AI Paper Atlas

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

Domain-Adversarial Training of Neural Networks

Yaroslav Ganin, Evgeniya Ustinova, Hana Ajakan, Pascal Germain, Hugo Larochelle, François Laviolette, Mario March, Victor Lempitsky

JMLR 2016 · Label-efficient and robust learning › Domain adaptation and generalization · approach: Adversarial domain adaptation · 8,507 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 domain-adversarial regularizer to a shallow NN (ablation)one-hidden-layer NN · Amazon reviews (original data)Poisson binomial win probability: .87Improved
use DANN instead of linear SVMDANN · Amazon reviews (original data)Poisson binomial win probability: .83Improved
run DANN on mSDA representations instead of NNDANN on mSDA · Amazon reviews (mSDA representation)Poisson binomial win probability: .92Improved
run DANN on mSDA representations instead of SVMDANN on mSDA · Amazon reviews (mSDA representation)Poisson binomial win probability: .88Improved
add domain-adversarial training to CNN (ablation)source-only CNN · MNIST to MNIST-Maccuracy: .5225 -> .7666Improved
add domain-adversarial training to CNN (ablation)source-only CNN · Syn Numbers to SVHNaccuracy: .8674 -> .9109Improved
add domain-adversarial training to CNN (ablation)source-only CNN · SVHN to MNISTaccuracy: .5490 -> .7385Improved
add domain-adversarial training to CNN (ablation)source-only CNN · Syn Signs to GTSRBaccuracy: .7900 -> .8865Improved
replace DANN with subspace alignmentsource-only CNN · Syn Numbers to SVHNaccuracy: .8674 -> .8644Hurt
add domain-adversarial training (ablation)source-only CNN · MNIST to SVHNaccuracy: approximately 0.25 accuracyNo clear effect
add domain-adversarial training to AlexNet (ablation)source-only AlexNet · Office Amazon to Webcamaccuracy: .642 -> .730Improved
add domain-adversarial training to AlexNet (ablation)source-only AlexNet · Office DSLR to Webcamaccuracy: .961 -> .964Improved

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