AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

Learning Self-Consistency for Deepfake Detection

Tianchen Zhao, Xiang Xu, Mingze Xu, Hui Ding, Yuanjun Xiong, Wei Xia

ICCV 2021 · Image generation › Autoregressive and latent generation · approach: Generalizable fake image detection · 413 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 pairwise self-consistency loss (lambda=10 vs lambda=0) to I2G training (ablation)ResNet-34 with I2G · DFR/CD2/DFDC/DFDC-P avg (cross-dataset)AUC (%): 73.74 -> 82.83Improved
add consistency loss (lambda=10 vs 0) on DFDC (ablation)ResNet-34 with I2G · DFDC (cross-dataset)AUC (%): 51.72 -> 67.52Improved
raise lambda from 10 to 100 (ablation)PCL + I2G · DFR/CD2/DFDC/DFDC-P avg (cross-dataset)AUC (%): 82.83 -> 82.09Mixed
add I2G augmentation when training on DF (ablation)PCL · DFR/CD2/DFDC/DFDC-P avg (trained on DF)AUC (%): 79.19 -> 86.37Improved
add I2G augmentation when training on DFDC-P (ablation)PCL · DFR/CD2/DFDC/DFDC-P avg (trained on DFDC-P)AUC (%): 74.78 -> 86.35Improved
add I2G augmentation when training on DFDC-P (ablation)PCL · DFDC-P (trained on DFDC-P)AUC (%): 95.53 -> 94.38Hurt
use I2G only as pretraining (then finetune on DF) instead of joint training (ablation)PCL · DFR/CD2/DFDC/DFDC-P avg (trained on DF)AUC (%): 79.19 -> 84.89Improved
use joint training with I2G instead of I2G pretraining only (ablation)PCL · DFR/CD2/DFDC/DFDC-P avg (trained on DF)AUC (%): 84.89 -> 86.37Improved
patch size 16x16 versus 4x4 (ablation)PCL + I2G · FF++ (cross-dataset)AUC (%): 98.32 -> 99.11Improved
patch size 32x32 versus 16x16 (ablation)PCL + I2G · FF++ (cross-dataset)AUC (%): 99.11 -> 98.74Hurt
deepen backbone from ResNet-18 to ResNet-34 (ablation)PCL + I2G · DFR/CD2/DFDC/DFDC-P avg (cross-dataset)AUC (%): 75.74 -> 82.83Improved
deepen backbone from ResNet-34 to ResNet-50 (ablation)PCL + I2G · DFR/CD2/DFDC/DFDC-P avg (cross-dataset)AUC (%): 82.83 -> 83.90Improved

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