AI Paper Atlas

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

Group-Wise Dynamic Dropout Based on Latent Semantic Variations

Zhiwei Ke, Zhiwei Wen, Weicheng Xie, Yi Wang, Linlin Shen

AAAI 2020 · Adversarial robustness and security › Certified and empirical robustness · approach: Robust overfitting and regularization · 8 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 original dropout with FDD-1D on FC featuresResNet-18 dropout · CIFAR-100accuracy (%): 75.59 -> 76.81Improved
Replace original dropout with FDD-1DResNet-18 dropout · Fashion-MNISTaccuracy (%): 93.97 -> 94.33Improved
Replace original dropout with FDD-1DResNet-18 dropout · CIFAR-10accuracy (%): 94.24 -> 94.65Improved
Replace original dropout with FDD-1DResNet-18 dropout · SVHNaccuracy (%): 96.52 -> 96.85Improved
Replace spatial dropout with FDD-2D on feature mapsResNet-18 dropout · CIFAR-100accuracy (%): 75.13 -> 76.83Improved
Use FDD-3D instead of FDD-2D (3D PCA projection) (ablation)ResNet-18 dropout · CIFAR-100accuracy (%): 76.83 -> 76.53No clear effect
Correlate dropout probability positively instead of negatively with feature density (ablation)ResNet-18 dropout · Fashion-MNIST, CIFAR-10, CIFAR-100, SVHNaccuracy (%)Hurt
Correlate dropout probability negatively with feature density (ablation)ResNet-18 dropout · Fashion-MNIST, CIFAR-10, CIFAR-100, SVHNaccuracy (%)Improved
Add FDD-1D dropout under FGSM attackResNet-18 dropout · CIFAR-100adversarial accuracy (%): 15 -> 66.28Improved
Add FDD-1D dropout under FGSM attackResNet-18 dropout · Fashion-MNISTadversarial accuracy (%): 5.85 -> 39.62Improved
Use original random dropout instead of no dropout under attackResNet-18 dropout · CIFAR-100adversarial accuracy (%): 14.72 -> 15No clear effect
Use spatial dropout instead of no dropout under attackResNet-18 dropout · Fashion-MNISTadversarial accuracy (%): 11.95 -> 9.93No clear effect

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