AI Paper Atlas

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

An Inverse-Adversarial and Difficulty-Adaptive Robust Vision-Language Model

Ruobing Xu, Junhao Dong, Xiaohua Xie

ECCV 2026 · Adversarial robustness and security › Attacks on text and graphs · approach: Vision-language model robustness · 0 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 visual prompt tuning and text adapter (ablation)CLIP head-only fine-tuning · Caltech101 novelrobust accuracy: 42.78 -> 52.52Improved
add visual prompt tuning and text adapter (ablation)CLIP head-only fine-tuning · EuroSAT novelrobust accuracy: 9.36 -> 13.76Improved
use text adapter instead of text prompt tuning (ablation)CLIP · Caltech101 novelrobust accuracy: 44.27 -> 48.92Improved
use visual prompt tuning instead of visual adapter (ablation)CLIP · Caltech101 novelrobust accuracy: 43.30 -> 45.24Improved
combine text prompt tuning with visual adapter (ablation)CLIP · EuroSAT novelrobust accuracy: 10.28 -> 10.32No clear effect
add inverse-adversarial learning, expert adaptation and difficulty-adaptive fusion (ablation)CLIP · Caltech101 novelrobust accuracy: 39.52 -> 52.52Improved
add inverse-adversarial learning, expert adaptation and difficulty-adaptive fusion (ablation)CLIP · EuroSAT novelrobust accuracy: 8.25 -> 13.76Improved
use IADA-RVLM instead of FAP (ablation)robust few-shot CLIP · DTD novelharmonic mean: 23.06 -> 26.02Improved
use IADA-RVLM instead of FAProbust few-shot CLIP · EuroSAT novelharmonic mean: 16.94 -> 19.02Improved
use IADA-RVLM instead of AdvMaPLe under AutoAttackrobust few-shot CLIP · OxfordPets novel (AutoAttack)robust accuracy: 26.39 -> 26.02Hurt
increase attack budget from eps=2/255 to 4/255IADA-RVLM · Caltech101 novel (PGD)robust accuracy: 52.52 -> 51.01Hurt
increase scaling factor beta (ablation)IADA-RVLM · low vs high difficulty datasetsnatural and robust accuracyMixed

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