AI Paper Atlas

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

Scalable Verified Training for Provably Robust Image Classification

Sven Gowal, Krishnamurthy (Dj) Dvijotham, Robert Stanforth, Rudy Bunel, Chongli Qin, Jonathan Uesato, Relja Arandjelovic, Timothy Mann, Pushmeet Kohli

ICCV 2019 · Adversarial robustness and security › Certified and empirical robustness · approach: Neural network robustness verification · 161 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
use IBP verified training instead of Wong et al. at large radiusIBP · MNIST (eps=0.3)verified error rate: 26.92% -> 8.05%Improved
use IBP instead of Wong et al.IBP · CIFAR-10 (eps=8/255)verified error rate: 79.21% -> 67.96%Improved
use IBP instead of Wong et al. at small radiusIBP · CIFAR-10 (eps=2/255)verified error rate: 49.96% -> 49.98%No clear effect
use IBP instead of Madry et al. adversarial trainingIBP · CIFAR-10 (eps=8/255)PGD error rate: 75.95% -> 65.23%Improved
use IBP instead of Madry et al. adversarial trainingIBP · MNIST (eps=0.3)PGD error rate: 3.73% -> 6.12%Hurt
use IBP instead of Madry et al. at small CIFAR radiusIBP · CIFAR-10 (eps=2/255)PGD error rate: 42.01% -> 45.09%Hurt
use IBP instead of Madry et al. adversarial trainingIBP · CIFAR-10 (eps=8/255)test error: 20.33% -> 50.51%Hurt
use IBP instead of Wong et al. for training speedIBP · MNIST small modeltraining time per epoch: 2 minutes -> 3.5 secondsImproved
train with IBP loss on downscaled ImageNet instead of nominal trainingIBP · Downscaled ImageNet (eps=1/255)PGD error rate: 100.00% -> 90.88%Improved
train with IBP so the IBP bound itself is a tight estimateIBP · MNIST (eps=0.3)IBP-only vs MIP verified error: 8.21% -> 8.05%Improved
use cross-entropy instead of hinge or softplus loss on each specification (ablation)IBP · MNIST, CIFAR-10, SVHNverified accuracyImproved
elide the last linear layer with the specification (ablation)IBP · MNIST, CIFAR-10, SVHNverified accuracyImproved

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