SOTAVerified

Adversarial Purification

A class of adversarial defense methods that remove adversarial perturbations using a generative model.

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 51–65 of 65 papers

TitleStatusHype
Advancing Adversarial Robustness Through Adversarial Logit Update—0
ZeroPur: Succinct Training-Free Adversarial PurificationCode0
Adversarial Purification of Information MaskingCode0
Carefully Blending Adversarial Training, Purification, and Aggregation Improves Adversarial RobustnessCode0
Detecting and Defending Against Adversarial Attacks on Automatic Speech Recognition via Diffusion ModelsCode0
DiffSmooth: Certifiably Robust Learning via Diffusion Models and Local SmoothingCode0
Diffusion-based Adversarial Purification for Intrusion DetectionCode0
High-Frequency Anti-DreamBooth: Robust Defense against Personalized Image SynthesisCode0
Language Guided Adversarial PurificationCode0
LISArD: Learning Image Similarity to Defend Against Gray-box Adversarial AttacksCode0
PCLD: Point Cloud Layerwise Diffusion for Adversarial PurificationCode0
Pre-trained Multiple Latent Variable Generative Models are good defenders against Adversarial AttacksCode0
Random Sampling for Diffusion-based Adversarial PurificationCode0
Robust Overfitting Does Matter: Test-Time Adversarial Purification With FGSMCode0
VideoPure: Diffusion-based Adversarial Purification for Video RecognitionCode0
Show:102550
← PrevPage 2 of 2Next →

No leaderboard results yet.