AI Paper Atlas

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

Admix: Enhancing the Transferability of Adversarial Attacks

Xiaosen Wang, Xuanran He, Jingdong Wang, Kun He

ICCV 2021 · Adversarial robustness and security › Black-box and transfer attacks · approach: Transferable adversarial examples · 325 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
admix a small portion of other-category images into the input gradient (Admix instead of SIM)MI-FGSM with SIM · Inc-v3 to Inc-v4attack success rate (%): 69.4 -> 82.6Improved
admix instead of SIMMI-FGSM with SIM · Inc-v3 to Inc-v3_ens3attack success rate (%): 32.5 -> 39.0Improved
replace SIM by Admix in SI-TIMSI-TIM · Inc-v3 to Inc-v3_ens3attack success rate (%): 48.2 -> 59.1Improved
replace SIM by Admix in SI-DIMSI-DIM · Inc-v3 to Inc-v3_ens3attack success rate (%): 48.0 -> 52.2Improved
replace SIM by Admix in SI-TI-DIM, ensemble attackSI-TI-DIM · IncRes-v2_ensattack success rate (%): 86.8 -> 90.0Improved
replace SI-TI-DIM by Admix-TI-DIM against advanced defensesSI-TI-DIM · Nine defense models averageattack success rate (%): 79.9 -> 83.3Improved
use Admix instead of SIM in ensemble attackSIM · IncRes-v2_ensattack success rate (%): 59.5 -> 67.8Improved
increase number of sampled images m2 from 0 to 3 (ablation)Admix · Inc-v3 source, six other modelsattack success rate (%)Improved
increase m2 beyond 3 (ablation)Admix · Inc-v3 source, normally trained modelsattack success rate (%)Hurt
increase m2 beyond 3 (ablation)Admix · Inc-v3 source, adversarially trained modelsattack success rate (%)Improved
increase admix strength eta from 0 to 0.2 (ablation)Admix · Inc-v3 sourceattack success rate (%)Improved
mix labels in mixup (Mixup vs Mixup without label mixing) (ablation)Mixup attack · Inc-v3 sourceattack success rate (%)Hurt

Later papers that compared with this method

10 later accepted papers report Admix as a baseline in their results tables; 10 of them report a better number on most of the results they share with it.

2 more not listed.

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