AI Paper Atlas

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

GradMDM: Adversarial Attack on Dynamic Networks

Jianhong Pan, Lin Geng Foo, Qichen Zheng, Zhipeng Fan, Hossein Rahmani, Qiuhong Ke, Jun Liu

TPAMI 2023 · Adversarial robustness and security › Certified and empirical robustness · approach: Attack evaluation and defense critique · 11 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 Power Loss and Complexity Gradient Masking (GradMDM vs ILFO)ILFO · ImageNet SkipNet+HRLARP (%): 49.6 -> 73.4Improved
GradMDM vs ILFOILFO · CIFAR-10 SkipNet-HRLARP (%): 5.2 -> 30.0Improved
GradMDM vs ILFO on dynamic width networkILFO · ManiDP ImageNetARP (%): 85.6 -> 99.3Improved
GradMDM vs ILFO on transformers and early-exit CNNILFO · DVT ImageNetARP (%): 85.0 -> 98.3Improved
add Complexity Gradient Masking (ablation)Baseline Loss · ImageNet SkipNet+HRLARP (%): 49.6 -> 54.0Improved
add Power Loss (ablation)Baseline Loss · ImageNet SkipNet+HRLARP (%): 49.6 -> 69.0Improved
combine Power Loss with CGM (ablation)Power Loss · ImageNet SkipNet+HRLARP (%): 69.0 -> 73.4Improved
increase gamma (imperceptibility weight) from 1e+0 to 1e+4 (ablation)GradMDM · ImageNet SkipNet+HRLARP (%) and PSNR: 76.1 -> 57.5 ARP, 54.3 -> 73.9 PSNRMixed
increase Power Loss exponent alpha from 1 to 4 (ablation)GradMDM · ImageNet SkipNet+HRLARP (%): 50.7 -> 73.4Improved
increase alpha beyond 4 (ablation)GradMDM · ImageNet SkipNet+HRLARP (%): 73.4 -> 72.5 (alpha 5), 69.0 (alpha 6)No clear effect
update with Finished Gradient jointly instead of masking (ablation)GradMDM · ImageNet SkipNet+HRLARP (%): 73.4 -> 38.1Hurt
project out conflicting direction via CGM instead of adding L_F gradient (ablation)GradMDM · ImageNet SkipNet+HRLARP (%): 66.8 -> 73.4Improved

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