AI Paper Atlas

Research map / Adversarial robustness and security / Attacks on text and graphs

Adversarial Robustification via Text-to-Image Diffusion Models

Daewon Choi, Jongheon Jeong, Huiwon Jang, Jinwoo Shin

ECCV 2024 · Adversarial robustness and security › Attacks on text and graphs · approach: Vision-language model robustness · 1 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 self-adaptation (T2I personalization and classifier fine-tuning) (ablation)CLIP + denoised smoothing · ImageNet, eps=0.5certified accuracy (%): 29.6 -> 34.2Improved
add self-adaptation (T2I personalization and classifier fine-tuning) (ablation)ResNet-50 + denoised smoothing · ImageNet ResNet-50, eps=1.0robust accuracy (%): 44.2 -> 47.8Improved
adapt text-to-image diffusion model only (without classifier adaptation) (ablation)CLIP + denoised smoothing · ImageNet, eps=0.5certified accuracy (%): 29.6 -> 31.8Improved
also adapt classifier on top of T2I adaptation (ablation)CLIP + denoised smoothing · ImageNet, eps=0.5certified accuracy (%): 31.8 -> 34.2Improved
increase timestep correction factor k above 1 for super-resolution denoiser (ablation)denoiser · ImageNet subset denoised accuracyCLIP accuracy of denoised imagesImproved
add classifier-guided regularization (lambda=0.01) in personalization (ablation)T2I personalization · ImageNet, sigma=0.5ACR: 0.358 -> 0.394Improved
increase reference set from 1-shot to 4-shot per class (ablation)self-adaptation · ImageNet, sigma=0.5ACR: 0.409 -> 0.428Improved
increase number of noise samples n from 25 to 400 (ablation)denoised smoothing inference · ImageNet, eps=1.0robust vs clean accuracy: +9.2% robust, -3.8% clean, 8.0x inference timeMixed
use data-free denoised smoothing instead of Mao et al. adversarial fine-tuning on ImageNetCLIP · EuroSAT, eps=1.0robust accuracy (%): 0.0 -> 28.2Improved

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