AI Paper Atlas

Research map / Adversarial robustness and security / Certified and empirical robustness

Wasserstein Introspective Neural Networks

Kwonjoon Lee, Weijian Xu, Fan Fan, Zhuowen Tu

CVPR 2018 · Adversarial robustness and security › Certified and empirical robustness · approach: Robust overfitting and regularization · 63 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
Replace INN's classifier loss with a Wasserstein loss and gradient penaltyIntrospective neural networks · CIFAR-10Inception score: 1.95 ± .01 -> 4.62 ± .05Improved
Use a single WINN classifier rather than INNg's 20-CNN configurationUnsupervised texture modeling · Texture imagesvisual synthesis quality: 20Improved
Use WINN-single rather than INNgUnsupervised face modeling · CelebAvisual synthesis quality: 12 CNNsImproved
Use 5 cascaded classifiers instead of a single classifier (ablation)WINN · CIFAR-10Inception score: 4.62 ± .05 -> 5.58 ± .05Improved
Add WINN training to a vanilla CNNVanilla CNN (4 layers) · MNIST testerror: 0.89 % -> 0.67%Improved
Add WINN training to ResNet-32ResNet-32 · MNIST testerror: 0.45% -> 0.48%Hurt
Add WINN training to ResNet-32ResNet-32 · SVHN testerror: 4.64% -> 4.50%Improved
Train with WINN rather than standard cross-entropyVanilla CNN (4 layers) · MNIST adversarial examplesadversarial error: 32.41% -> 7.99%Improved
Train with WINN rather than standard cross-entropyResNet-32 · MNIST adversarial examplesadversarial error: 11.28% -> 2.05%Improved
Train with WINN rather than standard cross-entropyResNet-32 · SVHN adversarial examplesadversarial error: 29.29% -> 19.53%Improved
Replace a standard GAN discriminator with an introspective WINN classifierResNet discriminator trained on CelebA · CelebAvisual synthesis qualityImproved

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