Research map / Video / Video generation
X2Face: A network for controlling face generation using images, audio, and pose codes
Olivia Wiles, A. Sophia Koepke, Andrew Zisserman
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 identity losses (curriculum stage II) after the L1-only stage, tested with 1 source frame (ablation) | X2Face · VoxCeleb test | L1 reconstruction error: 0.0632 -> 0.0630 | Improved |
| add identity losses (curriculum stage II) after the L1-only stage, tested with 3 source frames (ablation) | X2Face · VoxCeleb test | L1 reconstruction error: 0.0524 -> 0.0521 | Improved |
| use 3 source frames instead of 1 at test time (stage I model) (ablation) | X2Face · VoxCeleb test | L1 reconstruction error: 0.0632 -> 0.0524 | Improved |
| use 3 source frames instead of 1 at test time (stage II model) (ablation) | X2Face · VoxCeleb test | L1 reconstruction error: 0.0630 -> 0.0521 | Improved |
| self-supervised linear pose prediction from driving vector versus a supervised pose network | X2Face driving vector · AFLW test | mean MAE (degrees): 7.02 -> 9.36 | Hurt |
Closest papers by meaning
- ControlFace: Harnessing Facial Parametric Control for Face Rigging (CVPR 2025 · similarity 0.88)
- HeadGAN: One-Shot Neural Head Synthesis and Editing (ICCV 2021 · similarity 0.88)
- Controllable Image-to-Video Translation: A Case Study on Facial Expression Generation (AAAI 2019 · similarity 0.88)
- CONFIG: Controllable Neural Face Image Generation (ECCV 2020 · similarity 0.88)
- Vec2Face: Scaling Face Dataset Generation with Loosely Constrained Vectors (ICLR 2025 · similarity 0.87)
- CapHuman: Capture Your Moments in Parallel Universes (CVPR 2024 · similarity 0.86)
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.
