Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Few-Shot Head Swapping in the Wild
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 HeSer instead of Deepfacelab for head swapping | Head swapping · VoxCeleb2 test set user study | ID: 2.16 -> 4.12 | Helped |
| Use HeSer instead of FaceSwap for head swapping | Head swapping · VoxCeleb2 test set user study | Exp: 3.02 -> 3.01 | No clear effect |
| Use HeSer instead of FaceSwap for head swapping | Head swapping · VoxCeleb2 test set user study | Skin Color: 3.36 -> 2.61 | Hurt |
| Use HeSer instead of Deepfacelab for head swapping | Head swapping · VoxCeleb2 test set user study | Inpainting: 0.05 -> 4.95 | Helped |
| Use HeSer instead of Deepfacelab for head swapping | Head swapping · VoxCeleb2 test set user study | Holistic: 2.26 -> 3.82 | Helped |
| Use HeSer instead of FOMM | One-shot face reenactment · Face reenactment, one-shot | E_ID: 0.49 -> 0.53 | Hurt |
| Use HeSer instead of LPD | One-shot face reenactment · Face reenactment, one-shot | E_P: 0.063 -> 0.026 | Helped |
| Increase source inputs to 32 shots and fine-tune (ablation) | HeSer · Face reenactment | SSIM: 0.77 -> 0.89 | Helped |
| Increase source inputs to 32 shots and fine-tune (ablation) | HeSer · Face reenactment | E_ID: 0.53 -> 0.22 | Helped |
| Increase source inputs to 32 shots and fine-tune (ablation) | HeSer · Face reenactment | E_P: 0.026 -> 0.024 | Helped |
| Add subject-specific fine-tuning (ablation) | HeSer · Face reenactment | pose reconstruction error | Hurt |
| Increase the K-shot number (ablation) | HeSer Head2Head Aligner · Face reenactment | identity similarity | Helped |
Closest papers by meaning
- Zero-Shot Head Swapping in Real-World Scenarios (CVPR 2025 · similarity 0.89)
- FuseAnyPart: Diffusion-Driven Facial Parts Swapping via Multiple Reference Images (NeurIPS 2024 · similarity 0.84)
- MeGA: Hybrid Mesh-Gaussian Head Avatar for High-Fidelity Rendering and Head Editing (CVPR 2025 · similarity 0.83)
- HRAvatar: High-Quality and Relightable Gaussian Head Avatar (CVPR 2025 · similarity 0.83)
- LUCAS: Layered Universal Codec Avatars (CVPR 2025 · similarity 0.83)
- What to Preserve and What to Transfer: Faithful, Identity-Preserving Diffusion-based Hairstyle Transfer (AAAI 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.