AI Paper Atlas

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

Accelerating Certified Robustness Training via Knowledge Transfer

Pratik Vaishnavi, Kevin Eykholt, Amir Rahmati

NeurIPS 2022 · Adversarial robustness and security › Certified and empirical robustness · approach: Certified training and ensembles · 2 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
train student with knowledge transfer instead of SmoothMixRegNet · CIFAR-10, sigma=0.25ACR: 0.538 -> 0.564Improved
train student with knowledge transfer instead of SmoothMixResNeXt · CIFAR-10, sigma=0.25ACR: 0.516 -> 0.558Improved
train student with knowledge transfer instead of SmoothMixDLA · CIFAR-10, sigma=0.25certified accuracy at radius 0.00: 77.72 -> 77.31No clear effect
use knowledge transfer instead of SmoothMix for training speedResNeXt/DLA/RegNet · CIFAR-10average per-epoch speedup: 10.75 ×Improved
use knowledge transfer instead of SmoothMix for total training timeResNeXt/DLA/RegNet · CIFAR-10total training time: 8.06 ×Improved
use knowledge transfer instead of SmoothMix for ResNeXtResNeXt · CIFAR-10total time (h): 45.21 -> 4.80Improved
apply knowledge transfer recursively (chain length 1 to 3) (ablation)KT framework · CIFAR-10, sigma=0.25ACR: 0.558 -> 0.559No clear effect
use a small ResNet-S teacher trained with SmoothMix to train ResNet-L (ablation)KT framework · CIFAR-10ACR: 0.550 -> 0.540No clear effect
use a small ResNet-S teacher to train ResNet-L (ablation)KT framework · CIFAR-10total time (h): 18.98 -> 10.07Improved
swap teacher training method from SmoothMix to MACER (ablation)KT framework · CIFAR-10 student ACRACR: 0.540 -> 0.528Hurt
swap teacher training method from SmoothMix to SmoothAdv (ablation)KT framework · CIFAR-10 student ACRACR: 0.540 -> 0.519Hurt
transfer robustness on ImageNet versus training student directlyKT framework · ImageNet, SmoothAdvACR: 0.820 -> 0.684Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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