SOTAVerified

Adversarial Attack

An Adversarial Attack is a technique to find a perturbation that changes the prediction of a machine learning model. The perturbation can be very small and imperceptible to human eyes.

Source: Recurrent Attention Model with Log-Polar Mapping is Robust against Adversarial Attacks

Papers

Showing 661670 of 1808 papers

TitleStatusHype
EvolBA: Evolutionary Boundary Attack under Hard-label Black Box condition0
Do we need entire training data for adversarial training?0
Analyzing the Noise Robustness of Deep Neural Networks0
Delving into Data: Effectively Substitute Training for Black-box Attack0
D-square-B: Deep Distribution Bound for Natural-looking Adversarial Attack0
DTA: Physical Camouflage Attacks using Differentiable Transformation Network0
Dual Teacher Knowledge Distillation with Domain Alignment for Face Anti-spoofing0
Applying Tensor Decomposition to image for Robustness against Adversarial Attack0
A Context-Aware Approach for Textual Adversarial Attack through Probability Difference Guided Beam Search0
Analyzing Sentiment Polarity Reduction in News Presentation through Contextual Perturbation and Large Language Models0
Show:102550
← PrevPage 67 of 181Next →

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1ResNet20Test Accuracy89.9589.95(1)Community Verified
2Xu et al.Attack: PGD2078.68Unverified
33-ensemble of multi-resolution self-ensemblesAttack: AutoAttack78.13Unverified
4TRADES-ANCRA/ResNet18Attack: AutoAttack59.7Unverified
5AdvTraining [madry2018]Attack: PGD2048.44Unverified
6TRADES [zhang2019b]Attack: PGD2045.9Unverified
7XU-NetRobust Accuracy1Unverified
#ModelMetricClaimedVerifiedStatus
13-ensemble of multi-resolution self-ensemblesAttack: AutoAttack51.28Unverified
2multi-resolution self-ensemblesAttack: AutoAttack47.85Unverified