Research map / Video / Video generation
Avatar Fingerprinting for Authorized Use of Synthetic Talking-Head Videos
Ekta Prashnani, Koki Nagano, Shalini De Mello, David P Luebke, Orazio Gallo
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 |
|---|---|---|---|
| use landmark pairwise-distance features instead of 3DMM features (ablation) | avatar fingerprinting network · NVFAIR test (face-vid2vid) | AUC: 0.718 -> 0.754 | Improved |
| replace ID-Reveal loss with the proposed contrastive embedding loss (ablation) | avatar fingerprinting network · NVFAIR test (face-vid2vid) | AUC: 0.754 -> 0.850 | Improved |
| add time-shuffled negative term R to the loss (ablation) | avatar fingerprinting network · NVFAIR test (face-vid2vid) | AUC: 0.850 -> 0.868 | Improved |
| increase clip duration from 31 to 71 frames (ablation) | avatar fingerprinting network · NVFAIR test (face-vid2vid) | AUC: 0.820 -> 0.868 | Improved |
| compress test videos heavily (CRF 5 to CRF 40) | avatar fingerprinting network · NVFAIR test (face-vid2vid) | AUC: 0.868 -> 0.849 | No clear effect |
| use our method instead of ID-Reveal or Agarwal et al. | avatar fingerprinting · NVFAIR test (face-vid2vid) | AUC: 0.720 / 0.726 -> 0.868 | Improved |
| train on one generator, test on unseen generators | avatar fingerprinting network · NVFAIR test (cross-generator) | AUC: average AUC 0.85 -> 0.83 | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- ID-Reveal — ID-Reveal: Identity-Aware DeepFake Video Detection (ICCV 2021) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Codec Avatar Studio: Paired Human Captures for Complete, Driveable, and Generalizable Avatars (NeurIPS 2024 · similarity 0.85)
- PEGASUS: Personalized Generative 3D Avatars with Composable Attributes (CVPR 2024 · similarity 0.84)
- GAIA: Zero-shot Talking Avatar Generation (ICLR 2024 · similarity 0.84)
- Make-Your-Anchor: A Diffusion-based 2D Avatar Generation Framework (CVPR 2024 · similarity 0.83)
- RAM-Avatar: Real-time Photo-Realistic Avatar from Monocular Videos with Full-body Control (CVPR 2024 · similarity 0.83)
- VLOGGER: Multimodal Diffusion for Embodied Avatar Synthesis (CVPR 2025 · similarity 0.83)
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.
