Research map / Probabilistic and scientific modeling / Molecules and proteins
Breaking Manifold Continuity: Vector Quantized Modeling for Real-Centric Deepfake Detection
Changshuo Wang, Jiangming Wang, Ke-Yue Zhang, Taiping Yao, Shouhong Ding, Ran Yi, Lizhuang Ma
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add discrete real codebook (ablation) | Effort baseline · DFDC and DFDCP avg. | video-level AUC: 0.876 -> 0.890 | Improved |
| add masked ViT (ablation) | codebook model · DFDC and DFDCP avg. | video-level AUC: 0.890 -> 0.894 | Improved |
| add adaptive tangent space projection (TSP) (ablation) | codebook + masked ViT · DFDC and DFDCP avg. | video-level AUC: 0.894 -> 0.900 | Improved |
| replace static averaging with hypernetwork-based fusion (ablation) | full model without fusion · DFDC and DFDCP avg. | video-level AUC: 0.900 -> 0.906 | Improved |
| add codebook on DFDC (ablation) | Effort baseline · DFDC | video-level AUC: 0.843 -> 0.861 | Improved |
| add all components on DFDCP (ablation) | Effort baseline · DFDCP | video-level AUC: 0.909 -> 0.932 | Improved |
| raise codebook size from 128 to 1024 (ablation) | codebook · FSh, DFDC, DFDCP avg. | video-level AUC: 0.860 -> 0.910 | Improved |
| raise codebook size from 1024 to 2048 (ablation) | codebook · FSh, DFDC, DFDCP avg. | video-level AUC: 0.910 -> 0.909 | No clear effect |
| raise codebook size from 1024 to 4096 (ablation) | codebook · FSh, DFDC, DFDCP avg. | video-level AUC: 0.910 -> 0.896 | Hurt |
| add full pipeline vs Effort baseline cross-dataset | Effort · Cross-dataset avg. | video-level AUC: 0.898 -> 0.929 | Improved |
| add full pipeline vs Effort baseline cross-method | Effort · Cross-method avg. | video-level AUC: 0.930 -> 0.954 | Improved |
| add modules to Effort in inference latency | Effort · Inference time per 100 batches | seconds: 16.24s -> 18.22s | Hurt |
Closest papers by meaning
- Transcending Forgery Specificity with Latent Space Augmentation for Generalizable Deepfake Detection (CVPR 2024 · similarity 0.87)
- Generalizing Deepfake Video Detection with Plug-and-Play: Video-Level Blending and Spatiotemporal Adapter Tuning (CVPR 2025 · similarity 0.86)
- Exploring Unbiased Deepfake Detection via Token-Level Shuffling and Mixing (AAAI 2025 · similarity 0.86)
- D^3: Scaling Up Deepfake Detection by Learning from Discrepancy (CVPR 2025 · similarity 0.85)
- Towards More General Video-based Deepfake Detection through Facial Component Guided Adaptation for Foundation Model (CVPR 2025 · similarity 0.85)
- Can We Leave Deepfake Data Behind in Training Deepfake Detector? (NeurIPS 2024 · similarity 0.85)
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.
