Research map / Image generation / Editing and style transfer
DynamicFace: High-Quality and Consistent Face Swapping for Image and Video using Composable 3D Facial Priors
Runqi Wang, Yang Chen, Sijie Xu, Tianyao He, Wei Zhu, Dejia Song, Nemo Chen, Xu Tang, Yao Hu
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 temporal motion module (AnimateDiff temporal attention) (ablation) | DynamicFace video · FF++ video | CLIP frame consistency: 95.78 -> 99.02 | Improved |
| add temporal motion module (ablation) | DynamicFace video · FF++ video | ID Retrieval: 99.21 -> 98.90 | Hurt |
| add temporal motion module (ablation) | DynamicFace video · FF++ video | Mouth error: 1.98 -> 2.30 | Hurt |
| add FusionTVO inference module (ablation) | DynamicFace with motion module · FF++ video | Aesthetic score: 3.48 -> 3.51 | Improved |
| add FusionTVO inference module (ablation) | DynamicFace with motion module · FF++ video | Consistency and warping error: 99.02 -> 99.04; 0.046 -> 0.045 | No clear effect |
| use both Face Former and ReferenceNet for identity injection (ablation) | DynamicFace · FF++ | ID Similarity: 0.520 -> 0.547 | Improved |
| use both Face Former and ReferenceNet for identity injection (ablation) | DynamicFace · FF++ | Expression error: 3.23 -> 2.32 | Improved |
| four composable facial conditions (background, normal, landmark, UV map) (ablation) | DynamicFace · qualitative ablation | visual quality | Improved |
| DynamicFace vs SimSwap on pose error | DynamicFace · FF++ | Pose error: 1.05 (SimSwap) vs 1.73 (Ours) | Hurt |
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.
- DiffSwap — DiffSwap: High-Fidelity and Controllable Face Swapping via 3D-Aware Masked Diffusion (CVPR 2023) · 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
- High-Fidelity Diffusion Face Swapping with ID-Constrained Facial Conditioning (CVPR 2026 · similarity 0.93)
- VividFace: A Robost and High-Fidelity Video Face Swapping Framework (NeurIPS 2025 · similarity 0.92)
- Controllable and Expressive One-Shot Video Head Swapping (ICCV 2025 · similarity 0.92)
- CanonSwap: High-Fidelity and Consistent Video Face Swapping via Canonical Space Modulation (ICCV 2025 · similarity 0.92)
- DiffSwap: High-Fidelity and Controllable Face Swapping via 3D-Aware Masked Diffusion (CVPR 2023 · similarity 0.92)
- OmniFace: Bridging the Image-to-Video Gap for High-Fidelity Face Swapping via Diffusion Transformer (ECCV 2026 · similarity 0.91)
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.
