AI Paper Atlas

Research map / Detection and segmentation / Open-set and few-shot detection

DeepfakeImpact: A Two-Stage Benchmark with Real-World Impact in Deepfake Detection

Chaoyu Gong, Han Zhang, Siqiang Luo

CVPR 2026 · Detection and segmentation › Open-set and few-shot detection · approach: Industrial anomaly detection

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
start from pre-trained backbone weights (ablation)SSTGNN (ResNet-50) · SMI-critical datasetAcc-SMI: 0.705 -> 0.741Improved
start from pre-trained backbone weights (ablation)FaceXray (ResNet-50) · SMI-critical datasetAcc-SMI: 0.247 -> 0.304Improved
start from pre-trained backbone weights (ablation)SPSL (Xception) · SMI-critical datasetAcc-SMI: 0.439 -> 0.462Improved
swap backbone ResNet-34 to ResNet-50 (ablation)SSTGNN · SMI-critical datasetAcc-SMI: 0.729 -> 0.741Improved
swap backbone among pre-trained ResNet-34/50, EfficientNet-B4, Xception (ablation)FaceXray, Core, SPSL · SMI-critical datasetAcc-SMI: FaceXray 0.304 for both ResNetsNo clear effect
add data augmentations (vs removing all) (ablation)frame-level detectors (XceptionNet, SIA, RECCE, Effort) · SMI-critical datasetrelative change in Acc-SMI: XceptionNet -2.20% without augmentationsImproved
add data augmentations (vs removing all) (ablation)video-level detectors (SSTGNN, STIL, FTCN, TALL) · SMI-critical datasetrelative change in Acc-SMI: SSTGNN +2.40%, STIL +4.50% when removedHurt
add Gaussian blur augmentation (ablation)frame-level detectors · SMI-critical datasetrelative change in Acc-SMI: removal gives +0.40% (SIA), +0.80% (XceptionNet)Hurt
add isotropic scaling augmentation (ablation)video-level detectors · SMI-critical datasetrelative change in Acc-SMI: removal gives -1.20% (STIL), -0.90% (SSTGNN)Improved
train with SMI-weighted cross-entropy instead of plain cross-entropy (ablation)SSTGNN · FF++ (t-SNE)feature separation (qualitative)Improved
use video-level temporal modeling instead of frame-level on face forgeryvideo-level vs frame-level baselines · Face forgery cross-domaintop-3 AUC: SSTGNN 0.8912 vs UIA-ViT 0.8040 (table)Improved
use video-level temporal modeling on AIG datavideo-level vs frame-level baselines · AIG datasets cross-domaintop-3 AUC: SSTGNN 0.8591 vs Effort 0.8583No clear effect

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