AI Paper Atlas

Research map / Image generation / Text-to-image diffusion

Efficient Input-level Backdoor Defense on Text-to-Image Synthesis via Neuron Activation Variation

Shengfang Zhai, Jiajun Li, Yue Liu, Huanran Chen, Zhihua Tian, Wenjie Qu, Qingni Shen, Ruoxi Jia, Yinpeng Dong, Jiaheng Zhang

ICCV 2025 · Image generation › Text-to-image diffusion · approach: Inference-time T2I improvement · 10 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
use neuron activation variation instead of raw neuron coverage (ablation)Ours · MS-COCO detectionAUROC: 54.5 -> 96.3Improved
use neuron activation variation instead of NC variation (ablation)Ours · MS-COCO detectionAUROC: 71.1 -> 96.3Improved
use all layers instead of only down blocks (ablation)Ours · Layer selectionAUROC: 88.0 -> 94.1Improved
use all layers instead of only mid blocks (ablation)Ours · Layer selectionAUROC: 90.9 -> 94.1Improved
use all layers instead of only attention layers (ablation)Ours · Layer selectionAUROC: 94.0 -> 94.1No clear effect
use all layers instead of only conv layers (ablation)Ours · Layer selectionAUROC: 93.5 -> 94.1No clear effect
use first-step activations instead of full 50-step generation (vs UFID)Ours · Detection timeseconds per sample: 33.041 -> 1.242Improved
compare with T2IShield on time costOurs · Detection timeseconds per sample: 7.445 -> 1.242Improved
compare against BadAct baselineOurs · T2I backdoor detectionAUC avg: 60.7 -> 95.3Improved
apply to Diffusion Transformer (SD v3.5) vs UFIDOurs · RickBKD-TPAAUROC: 48.3 -> 99.9Improved
reduce regularization weight alpha from 1e-5 to 1e-9 in adaptive attack (ablation)Adaptive attack · MS-COCOFID: 64.7 -> 13.2Mixed

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