AI Paper Atlas

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

Fastened CROWN: Tightened Neural Network Robustness Certificates

Zhaoyang Lyu, Ching-Yun Ko, Zhifeng Kong, Ngai Wong, Dahua Lin, Luca Daniel

AAAI 2020 · Adversarial robustness and security › Certified and empirical robustness · approach: Neural network robustness verification · 40 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
Optimize the activation bounding-line parameters used by CROWNCROWN · Sensorless Drive Diagnosis, l∞, 12 × [20] Sigmoidcertified l∞ bound: 0.07820402 -> 0.15097938Improved
Optimize the activation bounding-line parameters used by CROWNCROWN · MNIST, l1, 5 × [100] Sigmoidcertified l1 bound: 2.09984541 -> 2.51844215Improved
Optimize the activation bounding-line parameters used by CROWNCROWN · CIFAR10, l∞, 8 × [2048] Sigmoidcertified l∞ bound: 0.00237977 -> 0.00327576Improved
Optimize the activation bounding-line parameters used by CROWNCROWN · CIFAR10, l2, 6 × [2048] ReLUcertified l2 bound: 0.37454715 -> 0.37229005Hurt
Use LP-based bounds with multiple bounding lines instead of FROWNFROWN · Sensorless Drive Diagnosis, l∞, 4 × [20] ReLUcertified l∞ bound: 0.22465439 -> 0.22689147Improved
Use LP-based bounds with multiple bounding lines instead of FROWNFROWN · Sensorless Drive Diagnosis, l∞, 12 × [20] Sigmoidcertified l∞ bound: 0.15097938 -> 0.10809507Hurt

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