AI Paper Atlas

Research map / Adversarial robustness and security / Black-box and transfer attacks

Error-Correcting Output Codes with Ensemble Diversity for Robust Learning in Neural Networks

Yang Song, Qiyu Kang, Wee Peng Tay

AAAI 2021 · Adversarial robustness and security › Black-box and transfer attacks · approach: Robustness benchmarks and defenses · 10 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 ensemble diversity regularization (ablation)ECNN · CIFAR-10 PGD attackaccuracy (%): 17.0 -> 41.2Improved
Add parameter sharing at both ends of the encoder (ablation)ECNN^30_0.1 · CIFAR-10 PGD, ε=0.04accuracy (%): 51.1 -> 71.6Improved
Add parameter sharing at both ends of the encoder (ablation)ECNN^30_0.1 · CIFAR-10 normalaccuracy (%): 79.2 -> 85.1Improved
Increase ensemble size from 10 to 30 binary classifiers (ablation)ECNN^N_0.1 · CIFAR-10 PGD, ε=0.04accuracy (%): 41.2 -> 71.6Improved
Combine ECNN with adversarial training (ablation)ECNN^30_0.1 · CIFAR-10 PGD, ε=0.03accuracy (%): 59.6 -> 69.4Improved
Use ECNN rather than TanhEns16MNIST · MNIST PGD, ε=0.3accuracy (%): 79.2 -> 88.4Improved
Use ECNN rather than TanhEns16CIFAR-10 · CIFAR-10 PGD, ε=0.04accuracy (%): 63.1 -> 71.6Improved
Use ECNN rather than ResNet20CIFAR-100 · CIFAR-100 normalaccuracy (%): 62.2 -> 61.3Hurt
Use ECNN rather than ResNet20CIFAR-100 · CIFAR-100 PGD, ε=0.04accuracy (%): 5.3 -> 36.3Improved
Use ECNN rather than ResNet20GTSRB · GTSRB OptProjTranaccuracy (%): 48.2 -> 79.6Improved

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