AI Paper Atlas

Research map / Video / Video segmentation and tracking

Quality-Agnostic Deepfake Detection with Intra-model Collaborative Learning

Binh M. Le, Simon S. Woo

ICCV 2023 · Video › Video segmentation and tracking · approach: Deepfake video detection · 49 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
replace HSIC with soft-label collaborative loss (ablation)QAD (ResNet50) · NeuralTexturesAUC (%): 88.2 -> 84.0Hurt
add HSIC dependence maximization at intermediate layers (ablation)ResNet50 baseline · NeuralTexturesAUC (%): 88.2 -> 90.1Improved
use pairwise loss instead of HSIC (ablation)QAD (ResNet50) · NeuralTexturesAUC (%): 88.2 -> 89.1Mixed
add Adversarial Weight Perturbation (cross-entropy) (ablation)ResNet50 baseline · NeuralTexturesAUC (%): 88.2 -> 90.7Improved
use KL divergence instead of cross-entropy in AWP (ablation)QAD (ResNet50) · NeuralTexturesAUC (%): 90.7 -> 89.4Hurt
combine HSIC and AWP (ablation)ResNet50 baseline · NeuralTexturesAUC (%): 88.2 -> 91.3Improved
apply QAD training to different backbones (ablation)ResNet18/34, EfficientNet-B0 · 7 datasets avgAUC (%): +0.86 to +1.3 pointsImproved
one QAD model instead of separate per-quality modelsEfficientNet-B1 · Video compression, quality-awareAvg AUC (%): 96.39 -> 97.79Improved
QAD vs best baseline under random JPEG compression on NeuralTexturesQAD-E · NeuralTextures JPEGAUC (%): 79.00 -> 76.27Hurt
train QAD on FF++ and test cross-dataset on DFDCQAD-E · DFDCAUC (%): 71.9 -> 65.3Hurt
increase alpha above 0.002 (ablation)QAD (ResNet50) · NeuralTexturesAUCImproved

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